مدل‌سازی توان منابع آب زیرزمینی با استفاده از مدل های یادگیری ماشینی

سلامت, احمد; اردستانی, مجتبی; ملک محمدی, بهرام

doi:10.22092/wmrj.2023.360675.1506

مدل‌سازی توان منابع آب زیرزمینی با استفاده از مدل های یادگیری ماشینی

نوع مقاله : پژوهشی

نویسندگان

¹ دانشجوی دکتری مهندسی محیط‌زیست گرایش منابع آب، پردیس بین المللی کیش، دانشگاه تهران

² استاد، دانشکده ی محیط‌زیست، دانشگاه تهران

³ دانشیار، دانشکده ی محیط‌زیست، دانشگاه تهران

10.22092/wmrj.2023.360675.1506

چکیده

مقدمه و هدف
توانیابی منابع آب زیرزمینی یکی از اصل‌های پایه در مدیریت منابع آب است. هدف این پژوهش توان یابی آب زیرزمینی با استفاده از مدل های یادگیری ماشین بردار پشتیبان (SVM) و همچنین دستورالعمل های فراکاوشی (مدل ترکیبی (هیبریدی) ماشین بردار پشتیبان و دستورالعمل بهینه سازی فراکاوشی زنبورعسل (SVM-BA) و مدل ترکیبی ماشین بردار پشتیبان و دستورالعمل بهینه سازی فراکاوشی ازدحام ذرات (SVM-PSO) است.
مواد و روش ها
در این پژوهش در منطقه‌ی بجنورد عامل‌های بلندی، شیب، جهت، شاخص رطوبت پستی‌بلندی، فاصله از آبراهه، تراکم زهکشی، فاصله از گسل، سنگ شناسی، شاخص موقعیت پستی‌بلندی، شاخص ناهمواری زمین، موقعیت شیب نسبی و شاخص هم گرایی جریان انتخاب شدند. از شرکت آب منطقه ای اطلاعات موقعیت 359 چشمه دریافت شد. دستورالعمل تقسیم بندی تصادفی برای تقسیم نقطه‌های آموزشی (70%) و نقطه‌های اعتبارسنجی (30%) استفاده شد. براساس تحلیل حساسیت حذفی، اندازه‌ی اهمیت و مشارکت متغیرهای ورودی در توان یابی آب زیرزمینی مشخص شد. ارزیابی دقت مدل‌ها در دو مرحله‌ی آموزش و اعتبارسنجی براساس روش منحنی مشخصه عامل گیرنده (ROC) انجام شد.
نتایج
ارزیابی دقت مدل‌ها براساس معیار ارزیابی مساحت زیرمنحنی عامل گیرنده (AUC) نشان داد که دقت پیش بینی مدل ترکیبی ماشین بردار پشتیبان و دستورالعمل بهینه سازی فراکاوشی ازدحام ذرات (SVM-PSO) 945/0 بیشتر از دیگر مدل ها (SVM: 0.918 و SVM-BA: 0.932) بود. براساس نتایج مدل برتر در این پژوهش 7/75% از سطح منطقه طبقه‌ی توان زیاد و 38/66% از سطح منطقه طبقه‌ی توان خیلی‌زیاد را کسب کردند. از میان عامل‌های، موقعیت شیب نسبی (14/5%)، فاصله از گسل (13/4%) و سنگ شناسی (12/3%) در پیش بینی توان آب زیرزمینی بیشترین اهمیت را داشتند.
بحث و نتیجه گیری
براساس نتایج این پژوهش، عملکرد مدل ماشین بردار پشتیبان زیاد بود و دو دستورالعمل بهینه سازی فراکاوشی زنبورعسل و فراکاوشی ازدحام ذرات موجب تقویت قدرت پیشبینی مدل شدند. همچنین مدل های یادگیری ماشینی می توانند ارتباط میان عامل‌های محیطی و آب‌دهی چشمه ها را شناسایی کنند و با به‌کارگیری داده های موجود، نقش آن ها را تعیین کنند. عامل موقعیت شیب نسبی به‌عنوان مهم‌ترین متغیر و عامل فاصله از گسل نیز به‌عنوان دومین متغیر مهم در این پژوهش مشخص شدند. نتایج پژوهش نشان داد که در تغذیه‌ی بخش های زیرسطحی، ذخیره و جریان آب زیرزمینی، گسل های منطقه نقش مهمی داشتند. با کاربرد مدل در شناسایی وضعیت توان آب زیرزمینی عامل سنگ شناسی نیز به‌عنوان سومین متغیر مهم معرفی شد. در این پژوهش، با پیشنهاد نقشه‌ی توان آب زیرزمینی، امکان برنامه ریزی و تدقیق آمایش سرزمین برای آبخیز بجنورد فراهم شد.

کلیدواژه‌ها

عنوان مقاله [English]

Modeling Groundwater Potential Using Machine Learning Models

نویسندگان [English]

Ahmad Salamat ¹
Mojtaba Ardestani ²
Bahram Malekmohammadi ³

¹ Ph.D. Student in Environmental Engineering, Water resources, Kish International Campus, University of Tehran

² Professor, Faculty of Environment, University of Tehran

³ Associate Professor, Faculty of Environment, University of Tehran

چکیده [English]

Introduction
Finding the potential of groundwater resources is one of the basic principles in water resources management. The aim of this research is to determine the potential of groundwater using support vector machine learning (SVM) models as well as metaheuristic algorithms (hybrid support vector machine model and the bee metaheuristic optimization algorithm (SVM-BA) and hybrid model of the support vector machine and particle swarm optimization algorithm (SVM-PSO).
Materials and methods
The factors of elevation, slope, aspect, topographic humidity index, distance from stream, drainage density, distance from fault, lithology, topographic position index, land roughness index, relative slope position and flow convergence index were selected in Bojnurd region. Information on the location of 359 springs was received from the regional water company. Random division algorithm was used to divide training points (70%) and validation points (30%). Based on the removal sensitivity analysis, the importance and contribution of the input variables in determining the groundwater potential were determined. The accuracy of the models was evaluated in two stages of training and validation based on the receiver operating characteristic (ROC) curve method.
Results
The evaluation of the accuracy of the models based on the evaluation criteria of the area under curve (AUC) showed that the prediction accuracy of the hybrid model of the support vector machine and the particle swarm optimization algorithm (SVM-PSO) is 0.945 more than other models (SVM: 0.918 and SVM-BA: 0.932). Based on the results of the superior model, the high potential class and the very high potential class accounted for 7.75% and 38.66% of the area respectively. Among the factors, relative slope position with 14.5%, distance from the fault with 13.4% and lithology with 12.3% were the most important in predicting groundwater potential.
Discussion and Conclusion
Based on the results of this research, the support vector machine model has a high performance, and two optimization algorithms, the bee metaheuristic and particle swarm optimization algorithm, strengthen the predictive power of the model. Also machine learning models can identify the relationship between the environmental factors and the water supply of the springs and determine their role by using the available data. The relative slope position factor was identified as the most important variable and the distance from the fault factor was considered as the second most important variable in the present study. The results of the research showed that the faults in the region play an important role in aquifer recharge, storage and flow of groundwater. The lithological factor was also introduced by the model as the third important variable in identifying the state of groundwater potential. In this research, by presenting the groundwater potential map, it is possible to plan and verify land use planning for the Bojnurd watershed.

کلیدواژه‌ها [English]

Bojnurd
groundwater
machine learning
support vector machine

مراجع

Choubin B, Malekian A. 2017. Combined gamma and M-test based ANN and ARIMA models for groundwater fluctuation forecasting in semiarid regions. Environmental Earth Sciences, 76:538. https://doi.org/10.1007/s12665-017-6870-8.

Corsini A, Cervi F, Ronchetti F. 2009. Weight of evidence and artificial neural networks for potential groundwater spring mapping: An application to the Mt. Modino area (Northern Apennines, Italy). Geomorphology. 111(1-2): 79-87.

Dai FC, Lee CF, Li J, Xu ZW. 2001. Assessment of landslide susceptibility on the natural terrain of Lantau Island, Hong Kong. Environment Geology. 40(3): 381–391

Davoodi Moghaddam D, Rezaei M, Pourghasemi HR, Pourtaghie ZS, Pradhan B. 2013. Groundwater spring potential mapping using bivariate statistical model and GIS in the Taleghan Watershed, Iran. Arabian Journal Geoscience. 8: 913-929. https://doi.org/10.1007/s12517-013-1161-5.

De Santa Olalla FM, Dominguez A, Ortega F, Artigao A, Fabeiro C. 2007. Bayesian networks in planning a large aquifer in Eastern Mancha, Spain. Environmental Modelling and Software. 22(8):1089–1100.

Deepa S, Venkateswaran S, Ayyandurai R. 2016. Groundwater recharges potential zones mapping in upper Manimuktha Sub basin Vellar River Tamil Nadu India using GIS and remote sensing techniques. Modeling Earth Systems and Environment, 2: 137. https://doi.org/10.1007/s40808-016-0192-9.

Devkota KC, Regmi AD, Pourghasemi HR, Yoshida K, Pradhan B, Ryu IC. 2013. Landslide susceptibility mapping using certainty factor, index of entropy and logistic regression models in GIS and their comparison at Mugling-Narayanghat road section in Nepal Himalaya. Natural Hazards. 65(1):135–165.

Geroy IJ, Gribb MM, Marshall HP, Chandler DG, Benner SG, McNamara JP. 2011. Aspect influences on soil water retention and storage. Hydrology Processes. 25: 3836–3842.

Ghorbani Nejad S, Falah F, Daneshfar M, Haghizadeh A, Rahmati O. 2017. Delineation of groundwater potential zones using remote sensing and GIS-based data-driven models. Geocarto International. 32(2):167–187

Ghosh, S, Carranza EJM. 2010. Spatial analysis of mutual fault/fracture and slope controls on rock sliding in Darjeeling Himalaya, India. Geomorphology. 122: 1- 24.

Javhar A, Chen X, Jovid A, Yunus M, Jamshed A, Eldiiar D, Zulfiyor B. 2018. Evaluation of remote sensing techniques for lithological mapping in the southeastern pamir using landsat 8 OLI Data. International Journal of Geoinformatics. 14(1): 1-10.

Kalantar B, Pradhan B, Naghibi SA, Motevalli A, Mansor S. 2017. Assessment of the effects of training data selection on the landslide susceptibility mapping: A comparison between support vector machine (SVM), logistic regression (LR) and artificial neural networks (ANN). Geomatics, Natural Hazards and Risk. 9:(1): 49-69. DOI: 10.1080/19475705.2017.1407368.

Lee S, Kim JC, Jung HS, Lee MJ, Lee S. 2017. Spatial prediction of flood susceptibility using random-forest and boosted-tree models in Seoul metropolitan city, Korea. Geomatics, Natural Hazards and Risk. pp. 1-19.

Liu T, Yan H, Zhai L. 2015. Extract relevant features from DEM for groundwater potential mapping. The International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences. 40(7): 113-119.

Manap MA, Nampak H, Pradhan B, Lee S, Sulaiman WNA, Ramli MF. 2014. Application of probabilistic-based frequency ratio model in groundwater potential mapping using remote sensing data and GIS. Arabian Journal Geoscience. 7(2):711–724.

Mosavi A, Hosseini FS, Choubin B, Goodarzi M, Dineva AA, Sardooi ER. 2021. Ensemble boosting and bagging based machine learning models for groundwater potential prediction. Water Resources Management. 35(1):23-37.

Naghibi SA, Moghaddam DD, Kalantar B, Pradhan B, Kisi O. 2017. A comparative assessment of GIS-based data mining models and a novel ensemble model in groundwater well potential mapping. Journal of Hydrology. 548:471–483

Naghibi SA, Pourghasemi HR, Abbaspour K. 2018. A comparison between ten advanced and soft computing models for groundwater qanat potential assessment in Iran using R and GIS. Theor Appl Climatol. 131(3–4):967–984

Naghibi SA, Pourghasemi HR, Dixon B. 2016. GIS-based groundwater potential mapping using boosted regression tree, classification and regression tree, and random forest machine learning models in Iran. Environmental Monitoring and Assessment. 188(1): 44. https://doi.org/10.1007/s10661-015-5049-6.

Nampak H, Pradhan B, Manap MA. 2014. Application of GIS based data driven evidential belief function model to predict groundwater potential zonation. Journal of Hydrology. 513: 283-300.

Nhu V, Rahmati O, Falah F, Shojaei S, Ansari N, Shahabi H, Shirzadi A, Gorski K, Nguyen H, Ahmad B. 2020. Mapping of groundwater spring potential in karst aquifer system using novel ensemble bivariate and multivariate models. Water, 12(4):985; https://doi.org/10.3390/w12040985.

Ozdemir A. 2011. Using a binary logistic regression method and GIS for evaluating and mapping the groundwater spring potential in the Sultan Mountains (Aksehir, Turkey). Journal of Hydrology. 405(1–2):123–136.

Pham BT, Jaafari A, Prakash I, Singh SK, Quoc NK, Bui DT. 2019. Hybrid computational intelligence models for groundwater potential mapping. Catena. 182 p. https://doi.org/10.1016/j.catena.2019.104101.

Pham DT, Ghanbarzadeh A, Koç E, Otri S, Rahim S, Zaidi M. 2006. The bee’s algorithm—a novel tool for complex optimization problems. In Intelligent production machines and systems. Intelligent Production Machines and Systems. 2^ndI*PROMS Virtual International Conference 3–14 July 2006. pp. 454-459.

Phong T, Pham B, Trinh P, Ly H, Vu Q, Ho L, Phong L, Le H, Avand M, Prakash I. 2021. Groundwater potential mapping using GIS-based hybrid artificial intelligence methods. Groundwater. 59(5):745-760.

Pourghasemi HR, Moradi HR, Fatemi Aghda SM, Gokceoglu C, Pradhan B. 2013. GIS-based landslide susceptibility mapping with probabilistic likelihood ratio and spatial multi-criteria evaluation models (North of Tehran, Iran). Arabian Journal Geoscience. 7: 1857-1878. doi:10.1007/ s12517-012-0825-x.

Pourghasemi HR. Beheshtirad M. 2015. Assessment of a data-driven evidential belief function model and GIS for groundwater potential mapping in the Koohrang Watershed, Iran. Geocarto International. 30(6):662-685.

Rahmati O, Melesse AM. 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.

Rahmati O, Naghibi SA, Shahabi H, Bui DT, Pradhan B, Azareh A, Rafiei-Sardooi E, Samani AN, Melesse AM. 2018. Groundwater spring potential modelling: Comprising the capability and robustness of three different modeling approaches. Journal of Hydrology. 565:248-261.

Sachdeva S, Kumar B. 2021. Comparison of gradient boosted decision trees and random forest for groundwater potential mapping in Dholpur (Rajasthan), India. Stochastic Environmental Research and Risk Assessment. 35(2):287-306.

Shenga ZD, Baroková D, Šoltész A. 2018. Numerical modeling of groundwater to assess the impact of proposed rail way construction on groundwater regime. Pollack Periodica. 13(3): 187–196.

Tait NG, Davison RM, Leharne SA, Lerner DN. 2008. Borehole Optimization System (BOS): A case study assessing options for abstraction of urban groundwater in Nottingham, UK. Journal Environmental Modelling and Software. 23(5):611–621.

Thapa R, Gupta S, Guin S, Kaur H. 2017. Assessment of groundwater potential zones using multi-influencing factor (MIF) and GIS: a case study from Birbhum district, West Bengal. Applied Water Science. 7(7):4117-4131.

Thommeret N, Bailly JS, Puech C. 2009. Robust extraction of Thalwegs networks from DTMs for topological characterisation: a case study on badlands. Proceedings of Geomorphometry. 31: 218-223.

Yao X, Tham LG, Dai FC. 2008. Landslide susceptibility mapping based on support vector machine: A case study on natural slopes of Hong Kong, China. Geomorphology. 101(4):572-582.