Journal of Aquifer and Qanat
In collaboration Iranian Hydraulic Association

Comparison of numerical models and machine learning algorithms with groundwater level prediction approach: A case study in Birjand plainComparison of numerical models and machine learning algorithms with groundwater level prediction approach: A case

Document Type : Original Article

Authors

Mehran Malekinia 1
Amir Hossein Javid 2
Seyedeh Hoda Rahmati 3
Alireza Jahangir 4

1 Ph.D. Student, Department of Environmental Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran.

2 professor. Department of Natural Resources and Environment, SR.C., Islamic Azad University, Tehran, Iran

3 Assistant professor. Department of Natural Resources and Environment, SR.C., Islamic Azad University, Tehran, Iran

4 Assistant professor, Department of Water Science and Engineering, Payame Noor University, Tehran, Iran.

https://doi.org/10.22077/jaaq.2026.10568.1133
Abstract
با توجه به اهمیت تعیین حجم ذخایر منابع آب زیرزمینی برای مدیریت پایدار آب در مناطق خشک و نیمه‌خشک، این مطالعه به بررسی رویکردهای پیش‌بینی و شبیه‌سازی تراز آب زیرزمینی با مدل‌های یادگیری ماشین (LSTM، Random Forest و مدل ترکیبی Stacking Ensemble) در مقایسه با مدل عددی MODFLOW در 17 چاه مشاهده‌ای دشت بیرجند پرداخت. داده‌های سری زمانی از مارس 2008 تا جولای 2025 (204 نمونه) با معیارهای آماری تحلیل شدند. پیش‌پردازش داده‌ها شامل نرمال‌سازی و فیلتر صاف‌سازی برای حذف نویز بود. تحلیل حساسیت با روش‌های SHAP و مونت‌کارلو چاه‌های کلیدی را شناسایی کرد. مدل Stacking Ensemble، با تلفیق خروجی‌های کالیبره‌شده MODFLOW و پیش‌بینی‌های مدل‌های یادگیری ماشین، در چاه 1 بالاترین دقت را با R²=0.9783، RMSE=0.0322 متر و MAPE=0.0017% نشان داد، در حالی که MODFLOW با R²=0.968 و RMSE=3.17 متر در چاه‌های 5، 10 و 15 خطای بیشتری داشت. نمودارهای تیلور برتری مدل استکینگ را با ضریب همبستگی 0.9783 و انحراف معیار نرمال‌شده 0.812 نسبت به LSTM (ضریب همبستگی 0.9765، انحراف معیار 0.845) و Random Forest (ضریب همبستگی 0.9398، انحراف معیار 1.344) تأیید کردند. این رویکرد ترکیبی نیاز به داده‌های ورودی گسترده را کاهش می‌دهد، دقت پیش‌بینی را در آبخوان‌های ناهمگن با افت سالانه 40-60 سانتی‌متر بهبود می‌بخشد، و ابزاری کارآمد برای مدیریت پایدار منابع آب در دشت بیرجند و مناطق مشابه با شرایط هیدرولوژیکی پیچیده فراهم می‌کند.

Keywords

Keywords: Machine learning
Ensemble hybrid model
LSTM neural network
time series
Birjand plain
Subjects
Adombi, A. V. D. P., Chesnaux, R., Boucher, M.-A., Braun, M., & Lavoie, J. (2024). A causal physics-informed deep learning formulation for groundwater flow modeling and climate change effect analysis. Journal of Hydrology, 637, 131370. https://doi.org/10.1016/j.jhydrol.2024.131370
Aghlmand, R., & Abbasi, A. (2019). Application of MODFLOW with Boundary Conditions Analyses Based on Limited Available Observations: A Case Study of Birjand Plain in East Iran. Water, 11(9), 1904. https://doi.org/10.3390/w11091904
Alao, J. O., Bello, A., Lawal, H., & Abdullahi, D. (2024). Assessment of groundwater challenge and the sustainable management strategies. Results in Earth Sciences, 2, 100049. https://doi.org/10.1016/j.rines.2024.100049
Boo, K.B.W., El-Shafie, A., Othman, F., Khan, M.M.H., Birima, A.H., Ahmed, A.N., 2024. Groundwater level forecasting with machine learning models: A review. Water Research 252, 121249.DOI: https://doi.org/10.1016/j.watres.2024.121249
Borzì, I. (2025). Modeling Groundwater Resources in Data-Scarce Regions for Sustainable Management: Methodologies and Limits. Hydrology, 12(1), 11. https://doi.org/10.3390/hydrology12010011
Chang, Y.-W., Sun, W., Kow, P.-Y., Lee, M.-H., Chang, L.-C., & Chang, F.-J. (2025). Advanced groundwater level forecasting with hybrid deep learning model: Tackling water challenges in Taiwan’s largest alluvial fan. Journal of Hydrology, 655, 132887. https://doi.org/10.1016/j.jhydrol.2025.132887
Chen, C., Zhang, H., Shi, W., Zhang, W., & Xue, Y. (2023). A novel paradigm for integrating physics-based numerical and machine learning models: A case study of eco-hydrological model. Environmental Modelling & Software, 163, 105669. https://doi.org/10.1016/j.envsoft.2023.105669
Cui, X., Wang, Z., Xu, N., Wu, J., & Yao, Z. (2024). A secondary modal decomposition ensemble deep learning model for groundwater level prediction using multi-data. Environmental Modelling & Software, 175, 105969. https://doi.org/10.1016/j.envsoft.2024.105969
Di Salvo, C. (2022). Improving Results of Existing Groundwater Numerical Models Using Machine Learning Techniques: A Review. Water, 14(15), 2307. https://doi.org/10.3390/w14152307
Dunnington, D. W., Trueman, B. F., Raseman, W. J., Anderson, L. E., & Gagnon, G. A. (2021). Comparing the Predictive Performance, Interpretability, and Accessibility of Machine Learning and Physically Based Models for Water Treatment. ACS ES&T Engineering, 1(3), 348–356. https://doi.org/10.1021/acsestengg.0c00053
Ehteram, M., & Ghanbari-Adivi, E. (2023). Self-attention (SA) temporal convolutional network (SATCN)-long short-term memory neural network (SATCN-LSTM): An advanced python code for predicting groundwater level. Environmental Science and Pollution Research, 30(40), 92903–92921. https://doi.org/10.1007/s11356-023-28771-8
Ghebrehiwot, A. A., & Kozlov, D. V. (2019). Hydrological modelling for ungauged basins of arid and semi-arid regions: Review. Vestnik MGSU, 8, 1023–1036. https://doi.org/10.22227/1997-0935.2019.8.1023-1036
Gupta, H. V., Kling, H., Yilmaz, K. K., & Martinez, G. F. (2009). Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling. Journal of Hydrology, 377(1-2), 80–91. https://doi.org/10.1016/j.jhydrol.2009.08.003
Igwebuike, N., Ajayi, M., Okolie, C., Kanyerere, T., & Halihan, T. (2025). Application of machine learning and deep learning for predicting groundwater levels in the West Coast Aquifer System, South Africa. Earth Science Informatics, 18(1), 6. https://doi.org/10.1007/s12145-024-01623-w
Janssen, J., Tootchi, A., & Ameli, A. A. (2025). Tackling water table depth modeling via machine learning: From proxy observations to verifiability. Advances in Water Resources, 201, 104955. https://doi.org/10.1016/j.advwatres.2025.104955
Khan, J., Lee, E., Balobaid, A.S., Kim, K., 2023. A Comprehensive Review of Conventional, Machine Learning, and Deep Learning Models for Groundwater Level (GWL) Forecasting. Applied Sciences 13(4), 2743.DOI: https://doi.org/10.3390/app13042743
Navarro-Farfán, M. D. M., García-Romero, L., Martínez-Cinco, M. A., Hernández-Hernández, M. A., & Sánchez-Quispe, S. T. (2024). Comparison between MODFLOW Groundwater Modeling with Traditional and Distributed Recharge. Hydrology, 11(1), 9. https://doi.org/10.3390/hydrology11010009
Porhemmat, J., Sedghi, H., Babazadeh, H., & Fotovat, M. (2019). Evaluation of WEAP-MODFLOW model as integrated water resources management model for sustainable development. Civil Engineering Infrastructures Journal, Online First. https://doi.org/10.22059/ceij.2019.260084.1495
Pourmorad, S., Kabolizade, M., & Dimuccio, L. A. (2024). Artificial Intelligence Advancements for Accurate Groundwater Level Modelling: An Updated Synthesis and Review. Applied Sciences, 14(16), 7358. https://doi.org/10.3390/app14167358
Rad, M., Abtahi, A., Berndtsson, R., McKnight, U. S., & Aminifar, A. (2024). Interpretable machine learning for predicting the fate and transport of pentachlorophenol in groundwater. Environmental Pollution, 345, 123449. https://doi.org/10.1016/j.envpol.2024.123449
Rammohan, B., Partheeban, P., Ranganathan, R., & Balaraman, S. (2024). Groundwater Quality Prediction and Analysis Using Machine Learning Models and Geospatial Technology. Sustainability, 16(22), 9848. https://doi.org/10.3390/su16229848
Sheikh Khozani, Z., Barzegari Banadkooki, F., Ehteram, M., Najah Ahmed, A., & El-Shafie, A. (2022). Combining autoregressive integrated moving average with Long Short-Term Memory neural network and optimisation algorithms for predicting ground water level. Journal of Cleaner Production, 348, 131224. https://doi.org/10.1016/j.jclepro.2022.131224
Torres-Martínez, J. A., Mahlknecht, J., Kumar, M., Loge, F. J., & Kaown, D. (2024). Advancing groundwater quality predictions: Machine learning challenges and solutions. Science of The Total Environment, 949, 174973. https://doi.org/10.1016/j.scitotenv.2024.174973
Zarei, E., Saleh, F. N., & Dalir, A. N. (2024). Comparing the hybrid-lumped-LSTM model with a semi-distributed model for improved hydrological modeling. Journal of Water and Climate Change, 15(8), 4099–4113. https://doi.org/10.2166/wcc.2024.269
Journal of Aquifer and Qanat
Volume 6, Issue 2
December 2025
Pages 257-281

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us