ارزیابی آسیب پذیری آب زیرزمینی آبخوان دهگلان با استفاده از روش دراستیک، شبکه عصبی موجک و الگوریتم ازدحام مرغ

نویسندگان

1 دانش‌آموخته کارشناسی‌ارشد، گروه جغرافیا، دانشگاه تبریز، تبریز، ایران

2 استادیار گروه زمین‌شناسی، دانشگاه ارومیه، ارومیه، ایران

3 استادیار گروه مهندسی آب، دانشگاه آزاد اسلامی، واحد سنندج، سنندج، ایران

چکیده

شرق استان کردستان دارای پتانسیل مناسبی برای ذخیره منابع آب زیرزمینی است. اما بدلیل افزایش بی­رویه در طول سال­های متمادی و کاهش بارندگی­ها حجم ذخایر آبخوان دهگلان واقع در آن کاهش چشم­گیری داشته است و دشت دهگلان جزو دشت­های ممنوعه طبقه­بندی شده است. در این پژوهش به بررسی انتقال و پخش آلودگی در آبخوان دشت دهگلان با استفاده از روش دراستیک پرداخته شد. همچنین جهت مقایسه این روش با روش­های هوشمند از روش شبکه عصبی موجک و الگوریتم ازدحام مرغ استفاده گردید. روش دراستیک یکی از روش­های هم­پوشانی است که با هفت پارامتر اصلی مؤثر شامل عمق آب زیرزمینی، تغذیة خالص، محیط خاک، توپوگرافی، محیط غیراشباع و هدایت هیدرولیکی نقشه­ی حساسیت تهیه می­شود. در روش­های هوشمند پارامترهای دراستیک به عنوان ورودی و شاخص دراستیک به عنوان خروجی به مدل­ها معرفی گردید. برای این منظور هفت لایة رستری با مقیاس 1:25000 در محیط GIS تهیه شد و بعد از رتبه­دهی و وزن­دهی شاخص دراستیک بین 71 تا 153 به­دست آمد. نتایج اجرای مدل شبکة عصبی موجک و الگوریتم ازدحام مرغ به چهار بخش تقسیم­بندی شد، براساس نتایج مشاهده گردید که نقشه­های آسیب­پذیری نشان داد که بخش­های شمال آبخوان دارای پتانسیل آلودگی متوسط و زیادی است و بایستی محافظت بیش­تری از این مناطق صورت گیرد. با توجه به معیارهای ارزیابی نقشه آسیب­پذیری بدست آمده با شبکه عصبی موجک نسبت به الگوریتم ازدحام مرغ عملکرد بهتری داشته است. بطوریکه براساس معیارهای ارزیابی شاخص دراستیک در شبکه عصبی موجک به ترتیب در بخش­های غربی، جنوبی، شرقی و مرکزی (R2=98/0 و 82/0RMSE=)، (R2=8/0 و 51/1RMSE=)، (R2=96/0 و 69/0RMSE=) و (R2=92/0 و 2/1RMSE=) بدست آمد. همچنین در الگوریتم ازدحام مرغ معیارهای ارزیابی به ترتیب در بخش­های غربی، جنوبی، شرقی و مرکزی (R2=8/0 و 51/4RMSE=)، (R2=88/0 و 38/5RMSE=)، (R2=66/0 و 31/4RMSE=) و (R2=84/0 و 01/6RMSE=) بدست آمد. توزیع نیترات با شاخص آسیب­پذیری در حالت بهینه، نتایج بهتری را در پیش­بینی مناطق آلوده داشته است.

کلیدواژه‌ها


عنوان مقاله [English]

The vulnerability assessment of Dehgolan aquifer ground water using drastic method and wavelet neural network

نویسندگان [English]

  • G. Valadi 1
  • E. Abbas Novinpour 2
  • M. Byzidi 3
1 M. Sc., (graduated), Dept. of Geography, Tabriz University, Tabriz, Iran
2 Assist. Prof., Dept. of Geology, Urmia University, Urmia, Iran
3 Assist. Prof., Dept. of Water Engineering, Sanandaj Branch, Islamic Azad University, Sanandaj, Iran
چکیده [English]

The East part of Kurdistan province has got considerable potentials for saving ground water. However, the reservoir volume of Dehgolan aquifer has significantly decreased due to excessive consumption and low precipitation during past two decades. As a result, it has been classified as a forbidden plain. In this research, the transfer and spread of pollution in Dehgolan plain aquifer was investigated using the drastic method. Wavelet neural network and Chicken Swarming Algorithm model were used to compare the study method with other smart methods. Drastic method is one of Sensitivity map overlay methods which is produced with seven main parameters including ground water depth, recharge, topography, soil environment, unsaturated environment and hydraulic conduction. In smart methods, drastic parameters and index were respectively defined as model inputs and outputs. Accordingly, 7 raster layers with a scale of 1:25000 were produced in Arc GIS, and after ranking and weighing, the drastic index was obtained between 71 and 153. The results indicate that northern parts of the aquifer have moderate and high pollution potentials, so they need more protections. Based on evaluation criteria the obtained vulnerability maps using neural network model had a better performance than Chicken Swarming Algorithm so that R2 and RMSE were obtained as (R2=0.98 and RMSE= 0.82), (R2=0.8 and RMSE= 1.51), (R2=0.96 and RMSE= 0.69), (R2=0.92 and RMSE= 1. 2) for western, southern, eastern and central parts respectively. Moreover, R2 and RMSE were obtained by Chicken Swarming Algorithm as (R2=0.8 and RMSE= 4.51), (R2=0.88 and RMSE= 5.38), (R2=0.66 and RMSE= 4.31), (R2=0.84 and RMSE= 6. 01) for western, southern, eastern and central parts respectively. Based on nitrate distribution data, vulnerability index had better results in predicting polluted parts.

کلیدواژه‌ها [English]

  • Drastic Method
  • Quality
  • Nitrate
  • Artificial Intelligence
Abbas Novinpour, E., Moghimi, H., Kaki, M (2022) Aquifer vulnerability based on classical methods and GIS-based fuzzy optimization method (case study: Chahardoli plain in Kurdistan province, Iran). Arabian Journal of Geosciences. 15(360): 2-15.
Abbasnovinpour, E., Masoudi, S., Asghari Moghadam, A (2018) Assessing Vulnerability Aquifer of Nazloochay Plain of Urmia, Using DRASTIC Model and Verification with Nitrate Concentration in GIS. New Findings in Applied Geology. 12(23): 92-103. (In Persian) doi.org/10.22084/NFAG.2017.12964.1241.
Al-Abadi Alaa, M., Al-Shamma’a Ayser, M., Aljabbari Mukdad, H (2014) A GIS-based DRASTIC model for assessing intrinsic groundwater vulnerability in northeastern Missan governorate, southern Iraq, Appl Water Sci. doi.org/10.1007/s13201-014-0221-7.
Aller, L., Bennett, T., Lehr, J. H., Rebecca, P., Hackett, G (1987) DRASTIC: A Standardized System for Evaluating Ground Water Pollution Potential Using Hydrogeologic Settings, EPA 600/2-87-035. U.S. Environmental Protection Agency, Ada, Oklahoma. EPA/600/2-85-018.
Barzegar, R., Asghari Moghaddam, A., Baghban, H (2015) A supervised committee machine artificial intelligent for improving DRASTIC method to assess groundwater contamination risk: a case study from Tabriz plain aquifer, Iran, Stoch Environ Res Risk Assess. doi.org/10.1007/s00477-015-1088-3.
Behzad, M., Asghari, K., Coppola, Jr., Emery, A (2010) Comparative Study of SVMs and ANNs in Aquifer Water Level Prediction. J. Comput. Civ. Eng. 24: 408-413.
Fijani, E., Nadiri, A. A., Asghari Moghaddam, A., Tsai, F., T-C., Dixon, B (2013) Optimization of DRASTIC Method by Supervised Committee Machine Artificial Intelligence to Assess Groundwater Vulnerability for Maragheh-Bonab Plain Aquifer, Iran, Journal of hydrology. 530: 89-100. doi.org/10.1016/j.jhydrol.2013.08.038.
Ghosh, R., Sutradhar, S., Mondal, P., Das, N (2021) Application of DRASTIC model for assessing groundwater vulnerability: a study on Birbhum district, West Bengal, India. Modelling Earth Systems and Environment. 7(7): 3. doi.org/10.1007/s40808-020-01047-7.
Jovanovic, N. Z., Adams, S., Thomas, A., Fey, M., Beekman, H. E., Campbell, R., Saayman, I., Conrad, J (2006) Improved DRASTIC method for assessment of groundwater vulnerability to generic aqueous phase contaminants, WIT Transactions on Ecology and the Environment, 92, waste Management and the Environment III, 393–402. doi.org/10.2495/WM060421.
Kaveh, A., Iran Mansh, A (2010) Artificial neural network in structure optimization, third edition. Publications of the Building and Housing Research Center. (In Persian) https://www.sid.ir/journal/issue/6620/fa.
Krogulec, E., Trzeciak, J (2017) DRASTIC assessment of groundwater vulnerability to pollution in the Vistula floodplain in central Poland. Hydrology Research, 48 (4): 1088–1099. doi.org/10.2166/nh.2016.259.
Maqsoom, A., Aslam, B., Khalil, U., Ghorbanzadeh, O (2020) A GIS-based DRASTIC Model and an Adjusted DRASTIC Model (DRASTICA) for Groundwater Susceptibility Assessment along the China–Pakistan Economic Corridor (CPEC) Route. International Journal of Geo-Infornation, 9(5): 332. doi.org/10.3390/ijgi9050332.
Mohammad Zahiri, F., Safari, B., Bagheri, Z., Sobhan Ardakani, S (2014) Investigation of nitrate and nitrite pollution in underground water resources of Dehgolan city. Journal of Man and Environment, 12(3): 1-11. (In Persian)
Nadiri, A., Jabraili, N., Qarakhani, M (2018) Comparison of the ability of different combined methods in assessing the vulnerability of groundwater in the Dasht-Qaroh-Dehgolan aquifer. Journal of Eco-Hydrology, 6(3): 821-836. (In Persian).
Nadiri, K., Nadiri, A., Asghari Moghadam, A., Kurd, M (2017) A new method for identifying and determining areas at risk of subsidence (case study: Selmas plain aquifer). Journal of Eco-Hydrology, 5(1): 85-97. (In Persian) doi.org/10.22059/ije.2017.233252.601.
Patel, P., Mehta, D., Sharma, N (2022) A review on the application of the DRASTIC method in the assessment of groundwater vulnerability. Water supply. ws2022126.
Qarakhani, M., Nadiri, A., Asghari Moghadam, A., Sadeghi Aghdam, F (2016) Drastic model optimization using vector machine and artificial neural network to assess aquifer vulnerability. Ardabil Plain Eco-Hydrology Journal, 2(3): 311-324. (In Persian).
Safavi, A. A., Romagnoli, J. A (1997) Application of wavelet-based neural networks to modelling and optimization of an experimental distillation column, (IFAC Journal of) Engineering Applications of Artificial Intelligence, 10(3): 301-313.
Vrba, J., Zoporozec, A (1994) Guidebook on mapping groundwater vulnerability. International Contributions to Hydrogeology.Verlag Heinz Heise GmbH and Co. KG.
Wang, D., Safavi, A. A., Romagnoli, J. A (2000) Wavelet-based adaptive robust M-estimator for non-linear system identification., AIChE Journal, 46(8): 1607-1615.
Wang, W., Ding, J (2003) Wavelet Network Model and Its Application to the Prediction of Hydrology. Nature and Science, 1(1): 67-71.
Zouache, D., Ould Arby, Y., Nouioua, F., Ben Abdelaziz, F (2019) Multi-objective chicken swarm optimization: A novel algorithm for solving multi-objective optimization problems. Computers & Industrial Engineering.