بررسی عملکرد مدل‌های داده‌کاوی در پیش‌بینی بارش و تحلیل وضعیت خشک‌سالی ایستگاه سینوپتیک بندرعباس

نوع مقاله : مقاله پژوهشی

نویسندگان

1 گروه مهندسی آب و محیط‌زیست، دانشکده مهندسی عمران، دانشگاه صنعتی شاهرود، شاهرود، ایران.

2 گروه مهندسی محیط‌زیست، دانشکده مهندسی آب و محیط‌زیست، دانشگاه شهید چمران اهواز، اهواز، ایران.

3 دانشکده مهندسی عمران- سازه‌های هیدرولیکی، معاون طرح و توسعه شرکت آب منطقه‌ای یزد، یزد، ایران.

4 دانشکده مهندسی عمران، دانشگاه آزاد اسلامی واحد علوم و تحقیقات یزد، یزد، ایران.

10.22059/jwim.2023.355159.1053

چکیده

استفاده از روش‌های مختلف داده‌کاوی در پیش‌بینی خشک‌سالی متداول است. با این‌حال، به‌طور عمده انتخاب مدل برتر بر مبنای دقت شبیه‌سازی صورت می‌گیرد. درحالی‌که در اغلب مطالعات به ویژگی‌های ساختاری مدل‌ها کم‌تر توجه شده است. در این مقاله کارایی مجموعه‌ای از متداول‌ترین مدل‌های داده‌کاوی شامل شبکه عصبی مصنوعی چندلایه پرسپترون (ANN-MLP)، شبکه عصبی با تابع پایه شعاعی (ANN-RBF)، درخت تصمیم رگرسیونی (CART)، مدل درختی (M5P) و ماشین بردار پشتیبان (SVM) جهت پیش‌بینی بارش یک سال بعد ایستگاه سینوپتیک بندر‌عباس ارزیابی شده و ویژگی‌های هر یک از آن‌ها تشریح می‌شود. واسنجی و صحت‌سنجی مدل‌ها با استفاده از داده‌های خام و میانگین متحرک سه ساله پارامترهای اقلیمی در بازه آماری 1347 تا 1396 انجام شد. عملکرد مدل‌ها با استفاده از پارامترهای آماری مختلف و نمودارهای مقایسه‌ای ارزیابی شد. نتایج نشان داد مدل‌های SVM و M5P به‌ترتیب با مقادیر RMSE برابر 93/7 و 31/8 میلی‌متر، MAE برابر 66/3 و 69/4 میلی‌متر و ضریب همبستگی 83/0 و 82/0 کارایی مطلوبی در پیش‌بینی بارش دارند. هم‌چنین، به‌استثنای مدل CART، تغییر در ابزار داده‌کاوی تفاوت هشت تا 11 درصدی در دقت تخمین‌ها ایجاد می‌کند؛ بنابراین انتخاب مدل مناسب‌تر باید بر مبنای سایر ویژگی‌های روش‌ها در کنار میزان دقت آن‌ها صورت پذیرد. به‌علاوه، بهره‌گیری از میانگین متحرک سه ساله به‌طور متوسط ضریب همبستگی را حدود 78 درصد افزایش و RMSE را حدود 63 درصد کاهش داده است. تحلیل وضعیت درازمدت خشک‌سالی نشان داد با افزایش طول دوره شاخص بارش استاندارد، میزان تفکیک سال‌های مرطوب و خشک مشخص‌تر می‌شود.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Investigating the Performance of Data Mining Models in Rainfall Forecasting and Drought Analysis of Bandar Abbas Synoptic Station

نویسندگان [English]

  • Emad Mahjoobi 1
  • Hamid Abdolabadi 2
  • Javad Mahjoobi 3
  • Ehsan Ghafoori 4
1 Department of Water and Environmental Engineering, Faculty of Civil Engineering, Shahrood University of Technology, Shahrood, Iran.
2 Department of Environmental Engineering, Faculty of Water and Environmental Engineering, Shahid Chamran University of Ahvaz, Ahvaz, Iran.
3 Faculty of Civil Engineering–Hydraulic Structures, Plan and Development Deputy, Yazd Regional Water Company, Yazd, Iran.
4 Faculty of Civil Engineering, Islamic Azad University, Science and Research Branch, Yazd, Iran.
چکیده [English]

It is common to use different data mining methods in drought prediction. However, the selection of the best model is mainly based on the accuracy of the simulation, while most of the studies do not mention the features of the models. In this paper, the performance of the most common data mining models, including Multilayer Perceptron Artificial Neural Network (ANN-MLP), Radial Base Function Neural Network (ANN-RBF), Regression Decision Tree (CART), Model Tree (M5P), and Support Vector Machine (SVM) is evaluated in order to predict monthly one year ahead rainfall at Bandar Abbas synoptic station and then the characteristics of each of them are described. Calibration and validation of the models were done using raw data and a three-year moving average of climatic parameters from 1347 to 1396. The performance of the models has been evaluated using different statistical indices and comparative diagrams. The results showed that the SVM and M5P models have good prediction performance with RMSE of 7.93 and 8.31 mm, the MAE of 3.66 and 4.69 mm, and the CC of 0.83 and 0.82, respectively. Also, with the exception of the CART, the change in the data mining tool makes an eight to 11 percent difference in the accuracy of the estimates. Therefore, the most appropriate model should be selected based on other characteristics of the methods besides their accuracy. In addition, using the three-year moving average of the input parameters has increased the correlation coefficient by about 78 percent and reduced the RMSE by about 63 percent. The analysis of the long-term drought situation showed that with the increase in the period of the standard precipitation index, the separation of wet and dry years becomes more specific.

کلیدواژه‌ها [English]

  • Artificial neural network
  • Decision tree
  • Standard precipitation index
  • Support vector machine
  1. Abadeh, M., & Khosroshahi, M. (2021). Assessment and drought monitoring using Standardized Precipitation (SPI) and Standardized Precipitation Evapotranspiration (SPEI) Indices in Hormozgan province., Iranian Journal of Rangeland and Desert Research, 28(4), 718-732. (In Persian).
  2. Anwar, M. T., Hadikurniawati, W., Winarno, E., & Widiyatmoko, W. (2020, December). Performance Comparison of Data Mining Techniques for Rain Prediction Models in Indonesia. In 2020 3rd International Seminar on Research of Information Technology and Intelligent Systems (ISRITI) (pp. 83-88). IEEE.
  3. Bachmair, S., Stahl, K., Collins, K., Hannaford, J., Acreman, M., Svoboda, M., ... & Overton, I. C. (2016). Drought indicators revisited: the need for a wider consideration of environment and society, Wiley Interdisciplinary Reviews: Water, 3(4), 516-536.
  4. Boser, B. E., Guyon, I. M., & Vapnik, V. N. (1992, July). A training algorithm for optimal margin classifiers, In Proceedings of the fifth annual workshop on Computational learning theory (pp. 144-152).
  5. Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (2017). Classification and regression trees, Routledge.
  6. Dehbozorgi, M., Malekian, A., & Ehsani, A.H. (2015). Evaluation the efficiency of using artificial neural networks in predicting meteorological droughts in north-west of Iran. Journal of Geographical Sciences, 15(36), 139-156. (In Persian).
  7. Deo, R. C., Kisi, O., & Singh, V. P. (2017). Drought forecasting in eastern Australia using multivariate adaptive regression spline, least square support vector machine and M5Tree model, Atmospheric Research, 184, 149-175.
  8. Farahmand, A., & AghaKouchak, A. (2015). A generalized framework for deriving nonparametric standardized drought indicators, Advances in Water Resources, 76, 140-145.
  9. Ghorbani, K., Salari Jazi, M., & Abdolhosseini, M. (2015). Feasibility study of the prediction of annual drought based on drought conditions in the spring season. Iranian Journal of Irrigation & Drainage9(4), 636-645. (In Persian).
  10. Gringorten, I. I. (1963). A plotting rule for extreme probability paper, Journal of Geophysical Research, 68(3), 813-814.
  11. Haykin, S. (2004). Neural networks: A comprehensive foundation, 2, 41.
  12. Jahani, B., & Mohammadi, B. (2019). A comparison between the application of empirical and ANN methods for estimation of daily global solar radiation in Iran, Theoretical and Applied Climatology, 137(1), 1257-1269.
  13. Karami, A. (2010). Estimation of the critical clearing time using MLP and RBF neural networks, European Transactions on Electrical Power, 20(2), 206-217.
  14. Khoshhal Dastjerdi, J., & Hosseini, S.M. (2010). Application of artificial neural network in climatic elements simulation and drought cycle predication (case study: Isfahan province). Geography and Environmental Planning, 21(3 (39)), 107-120. (In Persian).
  15. Loukas, A., Vasiliades, L., & Tzabiras, J. (2007). Evaluation of climate change on drought impulses in Thessaly, Greece, European Water, 17(18), 17-28.
  16. Luenberger, D. G., & Ye, Y. (1984). Linear and nonlinear programming, (Vol. 2). Reading, MA: Addison-wesley.
  17. Malik, A., Kumar, A., Rai, P., & Kuriqi, A. (2021). Prediction of multi-scalar standardized precipitation index by using artificial intelligence and regression models, Climate, 9(2), 28.
  18. Mitchell, T. M. (1997). Machine learning. 1997. Burr Ridge, IL: McGraw Hill, 45(37), 870-877.
  19. Mokhtar, A., Jalali, M., He, H., Al-Ansari, N., Elbeltagi, A., Alsafadi, K., ... & Rodrigo-Comino, J. (2021). Estimation of SPEI meteorological drought using machine learning algorithms, IEEE Access, 9, 65503-65523.
  20. Mokhtarzad, M., Eskandari, F., Jamshidi Vanjani, N., & Arabasadi, A. (2017). Drought forecasting by ANN, ANFIS, and SVM and comparison of the models, Environmental Earth Sciences, 76(21), 1-10.
  21. Mozafari, Gh., Shafie, Sh., & Taghizade, Z. (2016). Evaluate the performance regression decision tree model in predicting drought (case study: synoptic station in Sanandaj). Journal of Natural Environment Hazards, 4(6), 1-19. (In Persian).
  22. Murad, S. H., & Salih, Y. M. M. (2020). Comparable Investigation for Rainfall Forecasting using Different Data Mining Approaches in Sulaymaniyah City in Iraq. International Journal of Advances in Life Science and Technology, 4(1), 11-18.
  23. Nikbakht Shahbazi, A., Zahraie, B., & Nasseri, M. (2012). Seasonal meteorological drought prediction using support vector machine. Journal of Water and Wastewater23(2), 73-85. (In Persian).
  24. Nivedika, M., Meghwal, M., & PV, R. (2021). Forecasting Drought via Soft-Computation Multi-layer Perceptron Artificial Intelligence Model, International Research Journal on Advanced Science Hub, 3, 30-36.
  25. Quinlan, J.R. (1992) Learning with Continuous Classes, Proceedings of Australian Joint Conference on Artificial Intelligence, Hobart 16-18 November 1992, 343-348.
  26. Roodposhti, M. S., Safarrad, T., & Shahabi, H. (2017). Drought sensitivity mapping using two one-class support vector machine algorithms, Atmospheric Research, 193, 73-82.
  27. Smola, A. J. (1996). Regression estimation with support vector learning machines, Doctoral dissertation, Master’s thesis, Technische Universität München.
  28. Vapnik Vladimir, N. (2000). The nature of statistical learning theory, Second Edition. Springer.
  29. Wu, J. (2018, August). Co-Evolution Algorithm for Parameter Optimization of RBF Neural Networks for Rainfall-Runoff Forecasting, In International Conference on Intelligent Computing (pp. 195-206). Springer, Cham.
  30. Zainudin, S., Jasim, D. S., & Bakar, A. A. (2016). Comparative analysis of data mining techniques for Malaysian rainfall prediction. International Journal on Advanced Science Engineering Information Technology, 6(6), 1148-1153.
  31. Zhang, Y., Yang, H., Cui, H., & Chen, Q. (2020). Comparison of the ability of ARIMA, WNN and SVM models for drought forecasting in the Sanjiang Plain, China, Natural Resources Research, 29(2), 1447-1464.