ترکیب روش‌ استخراج مشخصه با ریزمقیاس نمایی آماری مبتنی بر ترکیب مدل‌های هوش مصنوعی

نوع مقاله : مقاله پژوهشی

نویسندگان

1 عمران، دانشکده عمران، دانشگاه تبریز، تبریز، ایران

2 مدیریت منابع آب، دانشکده عمران، دانشگاه تبریز، تبریز، ایران

چکیده

در این پژوهش از دو مدل گردش عمومی جو GCM) (ESM-BNU, ESM2-Can ))برای شبیه‌سازی بارش دوره آتی در شهر تبریز، استفاده شده است. مهم‌ترین ضعف مدل‌های GCM ،بزرگ بودن مقیاس مکانی متغیرهای اقلیمی شبیه‌سازی شده است که روش‌های مختلف ریزمقیاس نمایی درصدد رفع این نقیصه می‌باشند. در این مطالعه برای ریزمقیاس نمودن متغیرهای اقلیمی مدل‌های GCM ،از مدل‌های هوش مصنوعی شبکه عصبی مصنوعی ( ANN )و نروفازی ( ANFIS ،)بهره گرفته شده است. بدون شک اصلی‌ترین مرحله به هنگام استفاده از این مدل ها، انتخاب مناسب ترین ورودی از میان داده‌های بسیار متعدد ارائه شده توسط GCM ها می‌باشد. بنابراین در این مطالعه برای انتخاب پارامترهای ورودی مؤثر از روش های درخت تصمیم و تابع اطلاعات مشترک ( MI )استفاده شده است. هم چنین روش ترکیب مدل برای کاهش عدم قطعیت در ریزمقیاس نمایی و افزایش دقت پیش‌بینی استفاده شده است. در این پژوهش مقایسه نتایج روش‌های ریزمقیاس نمایی نشان داد که، مدل ترکیبی با موثرترین ورودی‌های تعیین شده با درخت تصمیم نتایج مناسب تری ارائه می‌دهد. به طوریکه در هر دو مدل GCM ،بهکارگیری مدل ترکیبی با پیش‌بینی کننده‌های مبتنی بر درخت تصمیم نسبت به مدل‌های ANN و ANFIS در ریزمقیاس نمای سبب افزایش %38%-10 DC در مدل‌سازی بارش می‌گردد. پیشبینی بارش ایستگاه سینوپتیک تبریز با مدل ترکیبی نشان داد که بارش دوره آتی (2060-2020 ) تحت سناریوهای 5.RCP4 و 5.RCP8 تا %40 %-30 کاهش می‌یابد..

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

The conjunction of the feature extraction method with AI-based ensemble statistical downscaling models

نویسندگان [English]

  • zahra Razzaghzadeh 1
  • vahid nourani 2
  • nazanin behfar 2
1 civil engineer, faculty of civil engineering, Tabriz university, Tabriz, Iran
2 water resource management, factuly of civil engineering, university of Tabriz, Tabriz,Iran
چکیده [English]

In this study, two general circulation models (GCMs) (Can-ESM2, BNU-ESM) were used to simulate the future precipitation of Tabriz city. The weakness of GCMs is the coarse resolution of climate variables in which the different methods of downscaling is about to solve this deficiency. In this study, the Artificial Intelligence (AI) models, i.e., Artificial Neural Network (ANN) and Adaptive neuro-fuzzy inference system (ANFIS), were used to statistically downscale the climate variables of GCMs. Without any doubt, the most important step during the use of these models is selecting the dominant inputs among huge large-scale GCM data. So in this study for the selection of dominant inputs, decision tree, and mutual information (MI) feature extraction methods were used. Also, the ensemble techniques were used to evaluate the efficiency of downscaling models and to decrease the uncertainties. A comparison of the result of downscaling models indicated that the ensemble technique (i.e., hybrid of ANN and ANFIS) with dominant inputs based on decision tree feature extraction methods presents better performance. In both GCMs, the application of the downscaling ensemble couple with dominant predictors based on a decision tree model in precipitation downscaling showed 10%-38% increase in DC in versus the individual ANN and ANFIS downscaling models. The projection precipitation of Tabriz synoptic station for future (2020-2060) by proposed ensemble AI-based model indicated 30%-40% precipitation decreases under RCP4.5 and RCP8.5 scenarios.

کلیدواژه‌ها [English]

  • General Circulation Models (GCMs)
  • Adaptive neuro-fuzzy inference system (ANFIS)
  • Artificial Neural Network (ANN)
  • Mutual Information (MI)
  • Statistical Downscaling
[1]M. Rezaei, M. Nahtani, A. Abkar, M. Rezaei, M. Mirkazehi Rigi, The survey of the efficiency of SDSM for predicting temperature parameters in two dry and superhero climates (Case study: Kerman and Bam). , Watershed Management Research,  (2013) 117-131.
[2]R. Le Roux, M. Katurji, P. Zawar-Reza, H. Quénol, A. Sturman, Comparison of statistical and dynamical downscaling results from the WRF model, Environmental Modelling & Software, 100 (2018) 67-73.
[3]B. Timbal, Z. Li, E. Fernandez, The Bureau of meteorology statistical downscaling model graphical user interface:
user manual and software documentation, Citeseer, 2008.
[4]R.L. Wilby, C.W. Dawson, E.M. Barrow, SDSM—a decision support tool for the assessment of regional climate change impacts, Environmental Modelling & Software, 17(2) (2002) 145-157.
[5]P. Almasi, S. Soltani, Assessment of the climate change impacts on flood frequency (case study: Bazoft Basin, Iran), Stochastic Environmental Research and Risk Assessment, 31(5) (2017) 1171-1182.
[6]K. Goubanova, V. Echevin, B. Dewitte, F. Codron, K. Takahashi, P. Terray, M. Vrac, Statistical downscaling of sea-surface wind over the Peru–Chile upwelling region: diagnosing the impact of climate change from the IPSLCM4 model, Climate Dynamics, 36(7-8) (2011) 13651378.
[7]J. Liu, S. Chen, L. Li, J. Li, Statistical Downscaling and Projection of Future Air Temperature Changes in Yunnan Province, China, Advances in Meteorology, 2017 (2017).
[8]A. Anandhi, V. Srinivas, D.N. Kumar, R.S. Nanjundiah, Role of predictors in downscaling surface temperature to river basin in India for IPCC SRES scenarios using support vector machine, International Journal of Climatology: A Journal of the Royal Meteorological Society, 29(4) (2009) 583-603.
[9]J.T. Schoof, S. Pryor, Downscaling temperature and precipitation: A comparison of regression-based methods and artificial neural networks, International Journal of Climatology, 21(7) (2001) 773-790.
[10]M. Devak, C. Dhanya, A. Gosain, Dynamic coupling of support vector machine and K-nearest neighbour for downscaling daily rainfall, Journal of Hydrology, 525 (2015) 286-301.
[11]R. Chadwick, E. Coppola, F. Giorgi, An artificial neural network technique for downscaling GCM outputs to RCM spatial scale, Nonlinear Processes in Geophysics, 18(6) (2011).
[12]H. Mahsafar, R. MAKNOUN, B. Saghafian, The impact of climate change on Urmia Lake water level, Iran-Water Resources Research, 7 (2011) 47-58.
[13]G.P. Zhang, Time series forecasting using a hybrid ARIMA and neural network model, Neurocomputing, 50 (2003) 159-175.
[14]A. Mishra, V. Desai, V. Singh, Drought forecasting using a hybrid stochastic and neural network model, Journal of Hydrologic Engineering, 12(6) (2007) 626-638.
[15]V. Nourani, Ö. Kisi, M. Komasi, Two hybrid artificial intelligence approaches for modeling rainfall–runoff process, Journal of Hydrology, 402(1-2) (2011) 41-59.
[16]M.R. Najafi, H. Moradkhani, S.A. Wherry, Statistical downscaling of precipitation using machine learning with optimal predictor selection, Journal of Hydrologic Engineering, 16(8) (2010) 650-664.
[17]R. Haji Hosseini, J. Yazdi, S. Golian, Downscaling GCMs by Artificial Neural Network (ANN). , in:  Iran's second national irrigation and drainage congress., 2016.
[18]Z. Razzaghzadeh, V. Nourani, ANN based statistical downscaling of GCM model for prediction of hydroclimatic parameters (Case study: Tabriz City). , in:
16th Iranian Hydraulic Conference, 2017.
[19]V. Nourani, A.H. Baghanam, J. Adamowski, O. Kisi, Applications of hybrid wavelet–Artificial Intelligence models in hydrology: A review, Journal of Hydrology, 514 (2014) 358-377.
[20]V. Nourani, M.T. Sattari, A. Molajou, Threshold-based hybrid data mining method for long-term maximum precipitation forecasting, Water Resources Management, .8562-5462 )7102( )9(13
[21]Z. Alihosseini, nvestigation and Analysis of Wind Effects on Wind Climate Features Case Study: East Azarbaijan Province, M.Se thesis, Tabriz university, 2010.
[22]J. Guo, H. Chen, C.-Y. Xu, S. Guo, J. Guo, Prediction of variability of precipitation in the Yangtze River Basin under the climate change conditions based on automated statistical downscaling, Stochastic Environmental Research and Risk Assessment, 26(2) (2012) 157-176.
[23]S. Beecham, M. Rashid, R.K. Chowdhury, Statistical downscaling of multi-site daily rainfall in a South Australian catchment using a Generalized Linear Model, International journal of climatology, 34(14) (2014) 36543670.
[24]P.R. Tiwari, S. Kar, U. Mohanty, S. Kumari, P. Sinha, A. Nair, S. Dey, Skill of precipitation prediction with GCMs over north India during winter season, International Journal of Climatology, 34(12) (2014) 3440-3455.
[25]R.L. Wilby, S. Charles, E. Zorita, B. Timbal, P. Whetton, L. Mearns, Guidelines for use of climate scenarios developed from statistical downscaling methods, Supporting material of the Intergovernmental Panel on Climate Change, available from the DDC of IPCC TGCIA, 27 (2004) -.
[26]S.-T. Chen, P.-S. Yu, Y.-H. Tang, Statistical downscaling of daily precipitation using support vector machines and multivariate analysis, Journal of hydrology, 385(1-4) (2010) 13-22.
[27]C.E. Shannon, A mathematical theory of communications I and II. , Bell LABs Technical Journal, 27 (1948) 379-423.
[28]V.P. Singh, Hydrologic synthesis using entropy theory, Journal of Hydrologic Engineering, 16(5) (2011) 421-433.
[29]H.H. Yang, S. Van Vuuren, S. Sharma, H. Hermansky, Relevance of time–frequency features for phonetic and speaker-channel classification, Speech communication, 31(1) (2000) 35-50.
[30]Z. Gao, B. Gu, J. Lin, Monomodal image registration using mutual information based methods, Image and Vision Computing, 26(2) (2008) 164-173.
[31]E. Frank, Y. Wang, S. Inglis, G. Holmes, I.H. Witten, Using model trees for classification, Machine learning, .67-36 )8991( )1(23
[32]J.R. Quinlan, Learning with continuous classes, in:  5th Australian joint conference on artificial intelligence, World Scientific, 1992, pp. 343-348.
[33]M. Pal, S. Deswal, M5 model tree based modelling of reference evapotranspiration, Hydrological Processes: An International Journal, 23(10) (2009) 1437-1443.
[34]H.R. Maier, G.C. Dandy, Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications, Environmental modelling & software, 15(1) (2000) 101-124. [35] S. Haykin, Neural networks (Computer science), MacMillan College Publishing Co, New York, 1994.
[36]J.-S. Jang, ANFIS: adaptive-network-based fuzzy inference system, IEEE transactions on systems, man, and cybernetics, 23(3) (1993) 665-685.
[37]J.-S.R. Jang, C.-T. Sun, E. Mizutani, Neuro-fuzzy and soft computing; a computational approach to learning and machine intelligence, Prentice Hall, 1997.
[38]E. Sharghi, V. Nourani, N. Behfar, Earthfill dam seepage analysis using ensemble artificial intelligence based modeling, Journal of Hydroinformatics, 20(5) (2018) 1071-1084.
[39]D.R. Legates, G.J. McCabe Jr, Evaluating the use of “goodness of fit” measures in hydrologic and hydroclimatic model validation, Water resources research, 35(1) (1999) 233-241.
[40]V. Nourani, M. Komasi, A. Mano, A multivariate ANNwavelet approach for rainfall–runoff modeling, Water resources management, 23(14) (2009) 2877.
[41]O. Kisi, Evapotranspiration modelling from climatic data using a neural computing technique, Hydrological Processes: An International Journal, 21(14) (2007) 19251934.
[42]Z. Levin, W.R. Cotton, Aerosol pollution impact on precipitation: a scientific review, Springer Science & Business Media, 2008.
[43]A. Gholampour, R. Nabizadeh, S. Naseri, M. Yunesian,
H. Taghipour, N. Rastkari, S. Nazmara, S. Faridi, A.H. Mahvi, Exposure and health impacts of outdoor particulate matter in two urban and industrialized area of Tabriz, Iran, Journal of Environmental Health Science and Engineering, 12(1) (2014) 27.
[44]H. Sanikhani, Y. Dinpajoh, S. Pour Yusef, S.Z. Ghavidel, B. Solati, The Impacts of Climate Change on Runoff in Watersheds (Case Study: Ajichay Watershed in East Azerbaijan Province, Iran), Journal of Water and Soil, 27(6) (2014) 1225-1234.
[45]N. Toorini, M.R. Hesami kermani, Climate Change Prediction Using Nero Fuzzy (Case Study:Tehran and Tabriz Stations), Sharif Journal of Civil Engineering, 30(1) (2014) 139-147.