پیش‌بینی جریان ورودی سد امیرکبیر با استفاده از الگوهای دورپیوند اقلیمی و مدل‌های یادگیری ماشین

نوع مقاله : مقاله پژوهشی

نویسندگان

1 گروه مهندسی آب، دانشکده فناوری کشاورزی، دانشکدگان کشاورزی و منابع طبیعی، دانشگاه تهران، تهران، ایران.

2 گروه مهندسی آبیاری و آبادانی، دانشکده کشاورزی، دانشکدگان کشاورزی و منابع طبیعی، دانشگاه تهران، ایران.

10.22059/jwim.2023.354198.1044

چکیده

تقاضا برای آب شیرین به‌طور فزاینده­‌ای در حال افزایش است، درحالی‌که منابع محدود آب، در معرض اضافه­ برداشت، آلودگی و تغییرات اقلیمی هستند که این موارد نیاز به بهبود مدیریت منابع آب را به‌منظور توزیع عادلانه و دست‌یابی به اهداف توسعه پایدار برجسته می‌کند. یک گزینه کم­ هزینه برای حمایت از استراتژی‌­های مدیریت بهتر آب، توسعه مدل­‌هایی با قابلیت پیش‌­بینی مقادیر آب دردسترس، به‌ویژه مقادیر مربوط به بارش و جریان رودخانه‌­هاست. تنوع اقلیمی و تغییرات آب و هوایی، یک فرض اساسی برای پیش­بینی‌­های هیدروکلیماتولوژیکی است. یکی از جنبه‌­های قابل ­توجه این موضوع، همبستگی بین پدیده­‌های بزرگ ­مقیاس جوی-اقیانوسی یا الگوهای دورپیوند با فرایندهای هیدرولوژیکی در مقیاس محلی است که این الگوها می‌­توانند بر جریان ورودی به سدها نیز اثرگذار باشند. در این مطالعه از سه مدل یادگیری ماشین شبکه عصبی مصنوعی، شبکه عصبی بیزین و سیستم استنتاج عصبی-فازی سازگار برای پیش­‌بینی جریان ورودی به سدها بهره گرفته شده­ است تا کارایی آن‌ها مورد ارزیابی قرار بگیرد. بدین منظور 12 سناریو متشکل از متغیرهای بارش، جریان ورودی به سد و نُه شاخص­ اقلیمی با تأخیر تا شش گام زمانی، طراحی شد تا تأثیر استفاده از الگوهای دورپیوند به‌عنوان متغیرهای پیش­‌بینی ­کننده جریان یک ماه بعد سد امیرکبیر، موردبررسی قرار بگیرد. تحلیل نتایج این پژوهش نشان داد که استفاده از شاخص Nino3.4 با یک­ گام زمانی تأخیر و هم‌چنین شاخص PDO با دو گام زمانی تأخیر، می‌­توانند باعث افزایش دقت مدل نسبت به سناریوهای که در آن‌ها تنها از متغیرهای ایستگاهی استفاده شده ­است، شوند. طبق نتایج، شاخص Nino3.4 مؤثرترین شاخص بر جریان ورودی به سد امیرکبیر شناخته ­شد و سناریویی که در آن از شاخص نام­برده به همراه داده­‌های بارش و جریان یک و دوماه قبل به‌عنوان ورودی استفاده شده بود، در هر سه مدل، بالاترین دقت را به ثبت رساند. هم‌چنین عملکرد مدل ANFIS برای سناریوی نام­برده (سناریوی 9)، با مقادیر RMSE و R2، به‌ترتیب معادل با 69/5 مترمکعب بر ثانیه و 79/0، نسبت به دو مدل ANN و BNN بهتر بود، به ‌طوری‌که مقدار شاخص R2 برای بهترین سناریوی متشکل از متغیرهای ایستگاهی (سناریوی 5)، به میزان 15/0 افزایش یافته و مقدار شاخص RMSE نیز به میزان 78/0 مترمکعب کاهش یافته­ است.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Amirkabir Dam Inflow Prediction Using Teleconnection Patterns and Machine Learning Models

نویسندگان [English]

  • Ehsan Vasheghani 1
  • Ali Massah Bavani 1
  • Abbas Roozbahani 1
  • Farhad Behzadi 1
  • Misagh Bidabadi 2
1 Department of Water Engineering, Faculty of Agricultural Technology, University College of Agriculture & Natural Resources, University of Tehran, Tehran, Iran.
2 Department of Irrigation and Reclamation Engineering, Faculty of Agriculture, University College of Agriculture & Natural Resources, University of Tehran, Tehran, Iran.
چکیده [English]

The demand for freshwater is increasing, while the limited water resources are subject to over-harvesting, pollution, and climate change, which require improving water resource management to distribute it equitably and achieve It highlights the goals of sustainable development. A low-cost option to support better water management strategies is to develop models capable of predicting available water amounts, especially amounts related to precipitation and river flow. Climatic diversity and climate changes are basic assumptions for hydro climatological predictions. One of the remarkable aspects of this issue is the correlation between large-scale atmospheric-oceanic phenomena or Teleconnection patterns with hydrological processes on a local scale, and these patterns can also affect the inflow to the dams. This study uses three machine learning models, an artificial neural network, a Bayesian neural network, and an adaptive neuro-fuzzy inference system to predict dam inflow and evaluate their efficiency. For this purpose, 12 scenarios consisting of rainfall variables, inflow to the dam, and nine climatic indicators with a delay of up to six-time steps were designed to investigate the effect of using long-term models as predictive variables of the flow one month later in Amirkabir Dam. to be placed The analysis of the results of this research showed that the use of the Nino3.4 index with one-time step delay as well as the PDO index with two-time step delays can increase the accuracy of the model compared to the scenarios in which only station variables are used. to be According to the results, the Nino 3.4 index was found to be the most effective index on the inflow to Amirkabir Dam, and the scenario in which the mentioned index along with the rainfall and flow data of one and two months before was used as input, in all three The model recorded the highest accuracy. Also, the performance of the ANFIS model for the mentioned scenario (scenario 9), with RMSE and R2 values, equal to 5.69 and 0.79 cubic meters per second, respectively, was better than the ANN and BNN models, so the value of the R2 index for the best scenario consisting of station variables (scenario 5), it increased by 0.15 and the value of RMSE index decreased by 0.78 cubic meters.

کلیدواژه‌ها [English]

  • ANFIS
  • ANN
  • BNN
  • Inflow Prediction
  • Teleconnection Patterns
  1. Banihabib, M. E., & Jamali, F. S. (2011). Comparison of Dynamic Artificial Neural Network and Multivariate Linear Regression Models for Inflow Forecasting Using Remote Sensing Data, 20(2), 173. (In Persian)
  2. Barnston, A. G., & Livezey, R. E. (1987). Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Monthly weather review115(6), 1083-1126.
  3. Hurrell, J. W. (1995). Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science269(5224), 676-679.
  4. Jones, P. D., Jónsson, T., & Wheeler, D. (1997). Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and southwest Iceland. International Journal of Climatology: A Journal of the Royal Meteorological Society17(13), 1433-1450.
  5. Saji, N. H., & Yamagata, T. J. C. R. (2003). Possible impacts of Indian Ocean dipole mode events on global climate. Climate Research25(2), 151-169.
  6. Mohammadi, K., Eslami, H. R., & Dayyani, D. S. (2005). Comparison of regression, ARIMA, and ANN models for reservoir inflow forecasting using snowmelt equivalent (a case study of Karaj). Journal of Agriculture-Science-Technology, 7, 17-30. (In Persian)
  7. Sohrabi, S., & Bozorg Haddad, O. (2007). The artificial neural network model in predicting the inflow to the reservoirs of dams, the fourth national conference of watershed science and engineering of Iran, watershed management, Karaj. (In Persian)
  8. Rubaai, A., Castro-Sitiriche, M. J., & Ofoli, A. R. (2008). Design and implementation of parallel fuzzy PID controller for high-performance brushless motor drives: an integrated environment for rapid control prototyping. IEEE Transactions on Industry Applications44(4), 1090-1098.
  9. Yarahmadi, D., & Azizi, Gh. (2008). Multivariate analysis of relationship between sesonal rainfall in iran with climate indices. Geographical Research Quarterly, 39(62), 161-174. (In Persian)
  10. Talebi, Z. (2012). The effect of Teleconnection pattern of the North Atlantic Oscillation (NAO) on reference evapotranspiration in the western regions of the Iran. Thesis. (In Persian)
  11. Farajzadeh, M., Ahmadi, M., Alijani, B., Qavidel Rahimi, Y., Mofidi, A., & Babaeian, I. (2013). Study on Variation of Major Teleconnection Patterns (MTP) associated with Iran’s Precipitation. Journal of Climate Research1392(15), 31-45. (In Persian)
  12. Beltram, L., & Carbonin, D. (2013). ENSO teleconnection patterns on large-scale water resources systems. Thesis.
  13. Kumar, S., Tiwari, M. K., Chatterjee, C., & Mishra, A. (2015). Reservoir inflow forecasting using ensemble models based on neural networks, wavelet analysis, and bootstrap method. Water resources management29(13), 4863-4883.
  14. Azizi, G., Chehreara, T., & Safarrad, T. (2014). Simultaneous Effects of NAO and SOI Phases on Iran’s Climate. Geography and Environmental Sustainability4(3), 43-56. (In Persian)
  15. Salehizade, A. A., Rahmanian, M., Farajzadeh, M., & Ayoubi, A. (2015). Analysis of temperature changes on electricity consumption in Fars Province. Mediterranean Journal of Social Sciences, 6(3 S2), 610.
  16. Block, P. (2016). Tailoring seasonal climate forecasts for hydropower operations. Meteorology and Energy Security: Simulations, Projections, and Management, 179.
  17. Babaee Fini, O., & Fattahi, E. (2015). Seasonal Prediction of Discharge Entering into Uremia Lake by Using Climatic Large Scale Signals. Geography and Development13(40), 109-124. (In Persian)
  18. Ahmadi, F., Radmanesh, F., & Mir Abbasi Najaf Abadi, R. (2016). Application of Bayesian Networks and Genetic Programming for Predicting Daily River Flow (Case Study: Barandoozchay River). Irrigation Sciences and Engineering39(4), 213-223. (In Persian)
  19. Sedighi, F., Vafakhah, M., & Javadi, M. R. (2016). Application of Artificial Neural Network for Snowmelt-Runoff (Case Study: Latyan Dam Watershed). Journal of Watershed management research. 6 (12):43-54. (In Persian)
  20. Khosravi, D., & Mesgari, E. (2016). Spatial Analysis of Relationship Between Teleconnection Patterns and Monthly Temperature of Northwest Iran. Geography and Territorial Spatial Arrangement6(21), 203-214. (In Persian)
  21. Misaghi, F. (2016). Forecasting of the Alavian Dam Inflow Water Using Optimized Adaptive Neuro-Fuzzy Inference System (OANFIS). Iranian Journal of Soil and Water Research, 47(3), 439-448. (In Persian)
  22. Ruigar, H., & Golian, S. (2016). Prediction of precipitation in Golestan dam watershed using climate signals. Theoretical and applied climatology123(3), 671-682.
  23. Mbuvha, R., Jonsson, M., Ehn, N., & Herman, P. (2017, November). Bayesian neural networks for one-hour ahead wind power forecasting. In 2017 IEEE 6th International Conference on Renewable Energy Research and Applications (ICRERA)(pp. 591-596). IEEE.
  24. Mohammadi, M., Karami, H., Farzin, S., & Farokhi, A. (2017). Prediction of Monthly Precipitation Based on Large-scale Climate Signals Using Intelligent Models and Multiple Linear Regression (Case Study: Semnan Synoptic Station). Iranian Journal of Ecohydrology4(1), 201-214. (In Persian)
  25. Saligheh, M., & Sayadi, F. (2017). Summer precipitation determinant factors of Iran's South-East. Natural Environment Change3(1), 59-70. (In Persian)
  26. Steirou, E., Gerlitz, L., Apel, H., & Merz, B. (2017). Links between large-scale circulation patterns and streamflow in Central Europe: A review. Journal of Hydrology549, 484-500.
  27. Yang, T., Asanjan, A. A., Welles, E., Gao, X., Sorooshian, S., & Liu, X. (2017). Developing reservoir monthly inflow forecasts using artificial intelligence and climate phenomenon information. Water Resources Research53(4), 2786-2812.
  28. Alessandro, G. (2018). Informing water reservoir operations with climate teleconnections. Thesis.
  29. Kim, K., Lee, S., & Jin, Y. (2018). Forecasting quarterly inflow to reservoirs combining a copula-based Bayesian network method with drought forecasting. Water10(2), 233.
  30. Banihabib, M. E., Ahmadian, A., & Valipour, M. (2018). Hybrid MARMA-NARX model for flow forecasting based on large-scale climate signals, sea-surface temperatures, and rainfall. Hydrology Research49(6), 1788-1803.
  31. Ahmadi, M., Salimi, S., Hosseini, S. A., Poorantiyosh, H., & Bayat, A. (2019). Iran's precipitation analysis using synoptic modeling of major teleconnection forces (MTF). Dynamics of Atmospheres and Oceans, 85, 41-56.
  32. Babaei, M., Moeini, R., & Ehsanzadeh, E. (2019). Artificial neural network and support vector machine models for inflow prediction of dam reservoir (Case study: Zayandehroud dam reservoir). Water Resources Management33(6), 2203-2218.
  33. Samadi, M., & Fathabadi, A. (2019). Application of Time Series, ANN, and SVM Models in Forecasting the Gorgan Dam Inflow Rate. Environment and Water Engineering4(4), 299-309. (In Persian)
  34. Esmaili, K., Gandomkar, A., & Khodagholi, M. (2020). Identifying the Trend of Temperature Changes in the South Iranian Coasts and its Relationship with Teleconnections. Physical Geography Quarterly13(49), 1-22. (In Persian)
  35. Sabziparvar, A., Firoozmand, Z., & Varshavian, V. (2020). The Impact of Teleconnection Phenomena on Shifting the Date of First Autumn and Last Spring Frost Events: Physical Geography Research Quarterly, 52(2), 295-311. (In Persian)
  36. Maryanaji, Z., Tapak, L., & Hamidi, O. (2019). Climatic and atmospheric indices teleconnection impact the characteristics of frost season in western Iran. Journal of Water and Climate Change10(2), 391-401.
  37. Noorbeh, P., Roozbahani, A., & Kardan Moghaddam, H. (2020). Annual and monthly dam inflow prediction using Bayesian networks. Water Resources Management34(9), 2933-2951.
  38. Rasouli, K., Nasri, B. R., Soleymani, A., Mahmood, T. H., Hori, M., & Haghighi, A. T. (2020). Forecast of streamflows to the Arctic Ocean by a Bayesian neural network model with snow cover and climate inputs. Hydrology Research51(3), 541-561.
  39. Wagena, M. B., Goering, D., Collick, A. S., Bock, E., Fuka, D. R., Buda, A., & Easton, Z. M. (2020). Comparison of short-term streamflow forecasting using stochastic time series, neural networks, process-based, and Bayesian models. Environmental Modelling & Software126, 104669.
  40. Wang, J., Wang, X., hui Lei, X., Wang, H., hua Zhang, X., jun You, J., feng Tan, Q., & lian Liu, X. (2020). Teleconnection analysis of monthly streamflow using ensemble empirical mode decomposition. Journal of Hydrology, 582, 124411.
  41. Zhang, X., Wang, H., Peng, A., Wang, W., Li, B., & Huang, X. (2020). Quantifying the uncertainties in data-driven models for reservoir inflow prediction. Water Resources Management34(4), 1479-1493.
  42. Lee, D., Kim, H., Jung, I., & Yoon, J. (2020). Monthly reservoir inflow forecasting for the dry period using teleconnection indices: a statistical ensemble approach. Applied Sciences, 10(10), 3470.
  43. Linh, N. T. T., Ruigar, H., Golian, S., Bawoke, G. T., Gupta, V., Rahman, K. U., Sankaran, A., & Pham, Q. B. (2021). Flood prediction based on climatic signals using wavelet neural network. Acta Geophysica, 69(4), 1413-1426.
  44. Latif, S. D., Ahmed, A. N., Sathiamurthy, E., Huang, Y. F., & El-Shafie, A. (2021). Evaluation of deep learning algorithm for inflow forecasting: a case study of Durian Tunggal Reservoir, Peninsular Malaysia. Natural Hazards109(1), 351-369.
  45. Panahi, F., Ehteram, M., Ahmed, A. N., Huang, Y. F., Mosavi, A., & El-Shafie, A. (2021). Streamflow prediction with large climate indices using several hybrid multilayer perceptrons and copula Bayesian model averaging. Ecological Indicators133, 108285.
  46. Behzadi, F., Javadi, S., Yousefi, H., & Moridi, A. (2022). Investigation and analysis of the effect of drought on groundwater aquifers in Iran (Case study: Shahrekord plain): Journal of Water and Irrigation Management, 12(2), 327-348. (In Persian)
  47. Chu, H., Bian, J., Lang, Q., Sun, X., & Wang, Z. (2022). Daily Groundwater Level Prediction and Uncertainty Using LSTM Coupled with PMI and Bootstrap Incorporating Teleconnection Patterns Information. Sustainability14(18), 11598.
  48. Helali, J., Ghaleni, M. M., Hosseini, S. A., Siraei, A. L., Saeidi, V., Safarpour, F., Mirzaei, M., & Lotfi, M. (2022). Assessment of machine learning model performance for seasonal precipitation simulation based on teleconnection indices in Iran. Arabian Journal of Geosciences, 15(15), 1-24.
  49. (2023). NOAA’s National Weather Service Climate Prediction Center. https://www.cpc.ncep.noaa.gov/data/teledoc/teleintro.shtml. Accessed 1/12/2023.
  50. (2023). NOAA’s National Weather Service Glossary. https://w1.weather.gov/glossary. Accessed 1/12/2023.