بررسی عملکرد رویکردهای مختلف هوش مصنوعی در ریزمقیاسنمایی دما (منطقه مورد مطالعه: استان اردبیل) | ||
| فصلنامه علوم محیطی | ||
| مقاله 13، دوره 20، شماره 4، 1401، صفحه 243-258 اصل مقاله (520.78 K) | ||
| نوع مقاله: مقاله پژوهشی | ||
| شناسه دیجیتال (DOI): 10.48308/envs.2022.35101 | ||
| نویسندگان | ||
| محمد حسین جهانگیر* ؛ سید محمد احسان عظیمی | ||
| گروه انرژی های نو و محیط زیست، دانشکده علوم و فنون نوین، دانشگاه تهران، تهران، ایران | ||
| چکیده | ||
| سابقه و هدف: پیش بینی بلند و کوتاه مدت آب و هوا از جمله چالش های بسیار مهم محققان آب و اقلیم بوده است. به منظور فائق آمدن بر این چالش، ابزارهای متعددی از جمله مدل های گردش عمومی جوی-اقیانوسی، سناریوهای پیش بینی و مدل های ریزمقیاس نمایی توسعه و استفاده شده است. این ابزارها با ایجاد همبستگی بین داده های بزرگ مقیاس مدل های گردش عمومی و داده های کوچک مقیاس سینوپتیک، به ریزمقیاس نمایی سناریوهای پیش بینی می کنند. مواد و روش ها: در این مطالعه پارامترهای پیش بینی کننده بزرگ مقیاس دوره آماری 1961 تا 2003 از پایگاه داده مراکز ملی پیش بینی محیط زیست (NCEP) ، داده های بزگ مقیاس سناریوهای پیش بینی A1B و A2 مدل HadCM3 دوره آماری 2001 تا 2100 از مرکز ارزیابی و مدلسازی اقلیم کانادا موسوم به CCCma، و داده های سینوپتیک هواشناسی ایستگاه های اردبیل از سازمان هواشناسی دریافت شده است. در این مطالعه سه روش (SDSM)، حداقل مربعات ماشین بردار پشتیبان (LS-SVM) و پرسپترون چند لایه (MLP) برای ریزمقیاس نمایی، و از سنجنده های آماری CC، MSE، RMSE، NMSE، Nash-Sutcliffe، MAE و دیاگرام تیلور به منظور ارزیابی کارایی روش های ریز مقیاس نمایی مورد استفاده قرار گرفت. نتایج و بحث: نتایج نشان داده است که مدل MLP بر اساس میانگین ایستگاه ها بهترین نتیجه را با مقادیر (CC=0/85)، (NMSE=0/63)، (NSH=0/73) و (MAE=0/52) کسب کرده و در رتبه های دوم و سوم به ترتیب مدل های LS-SVM و SDSM قرار گرفته اند. دیاگرام تیلور نیز مدل SDSM را ضعیف ترین مدل شناسایی و نتایج دو مدل LS-SVM و MLP را با اختلاف کمی با یکدیگر به عنوان مدل های برتر معرفی کرد. بر اساس نتایج ریزمقیاس نمایی، تمامی سناریوهای پیش بینی افزایش دمای میانگین روزانه تا سال 2100 در تمامی ایستگاه های مورد مطالعه را پیش بینی کرده اند. نتیجه گیری: بر اساس نتایج مطالعه، در تمامی سناریوهای پیش بینی و بر اساس تمامی روش های ریز مقیاس نمایی، افزایش دمای میانگین روزانه پیش بینی شده است، لذا لازم است تا در تهیه سیاست های توسعه ای و محیط زیستی این موضوع در نظر گرفته بشود. | ||
| کلیدواژهها | ||
| MLP؛ LS-SVM؛ SDSM؛ ریز مقیاس نمایی؛ دما میانگین روزانه؛ سناریوهای تغییر اقلیم؛ استان اردبیل | ||
| عنوان مقاله [English] | ||
| Evaluating the performance of artificial intelligence models for temperature downscaling (Study area: Ardabil province) | ||
| نویسندگان [English] | ||
| Mohammad Hossein Jahangir؛ Seyed Mohammad Ehsan Azimi | ||
| Department of Renewable Energies and Environment, Faculty of New Sciences and Technologies, University of Tehran, Tehran, Iran | ||
| چکیده [English] | ||
| Introduction: Long and short term weather forecasting have been as two of the most important challenges to water and climate researchers. In order to overcome this challenge, several tools, including atmospheric-ocean general circulation model forecasting scenarios, and downscaling models have been developed and used. These tools downscale forecasting scenarios by creating relationship between parameters of synoptic stations and Large-scale data of general circulation models. Material and methods: In this study large-scale predictor parameters from 1961 to 2003 from the database of National Centers for Environmental Prediction (NCEP), large-scale data for the A1B and A2 forecast scenarios of the HadCM3 model from 2001 to 2100 from the Canadian Centre for Climate Modelling and Analysis (CCCMA), and the meteorological synoptic data of Ardabil stations from the Meteorological Organization were gathered. To this end, three downscaling models such as Statistical Downscaling Model (SDSM), least squares support vector machine (LS-SVM) and multi-layer perceptron (MLP) were determined for downscaling; and correlation coefficient (CC), Mean Squared Error (MSE), Root Mean Square Error (RMSE), Normalized mean square error (NMSE), Nash-Sutcliffe, Mean Absolute Error (MAE), and Taylor diagram were used to evaluate the efficiency of the models. Results and discussion: The results showed that the MLP obtained the best results based on the average of the stations with the values of (CC=0.85), (NMSE=0.63), (NSH=0.73) and (MAE=0.52), and LS-SVM and SDSM are ranked second and third, respectively. Taylor's diagram also identified the SDSM as the weakest and the LS-SVM and MLP as superior models with a slight difference. Based on downscaling results of all forecasting scenarios, an increase in average daily temperature is also predicted by 2100 in all studied stations. Conclusion: Based on the results of the study, all forecasting scenarios and all methods of downscaling show increasing the daily average temperature by 2100. Hence, it is necessary to be taken this issue account for making environmental and developing policies in this area. | ||
| کلیدواژهها [English] | ||
| MLP, LS-SVM, SDSM, Downscaling, Average daily temperature, Climate change scenarios, Ardabil province | ||
| مراجع | ||
|
Ahmadi, B.N., Shirvani, A. and Nazemosadat, M., 2014. The Application of and for downscaling GCMS outputs for prediction of precipitation in across Southern Iran. Journal of water and soil (Agricultural sciences and technology). 28(5), 1037-1047. (In Persian with English abstract) Asakare, H., 2017. Comparing the performance of the SDSM models and those based on artificial neural networks in predicting the changes in minimum temperatures (station in case: urmia). The Journal of Spatial Planning. 21(4), pp.140-160. (In Persian with English abstract) Dehghani, N., Ghasemieh, H., Sadatinejad, S.J., Ghorbani, Kh. and Besalatpour, A.A., 2017. Comparative comparison of data mining models in downscaling rainfall and temperature (Case Study: Bazoft-e- Samsami Watershed). Journal of Water and Soil Conservation. 24(5), 227-240. (In Persian with English abstract) Dibike, Y.B. and Coulibaly, P., 2007. Validation of hydrological models for climate scenario simulation: the case of Saguenay watershed in Quebec. Hydrological Processes: An International Journal. 21(23), 3123-3135. Etemadi, H., Samadi, S.Z., Sharifikia, M. and Smoak, J.M., 2016. Assessment of climate change downscaling and non-stationarity on the spatial pattern of a mangrove ecosystem in an arid coastal region of southern Iran. Theoretical and Applied Climatology. 126(1-2), 35-49. Emami, F. and Koch, M., 2018. Evaluation of statistical-downscaling/bias-correction methods to predict hydrologic responses to climate change in the zarrine river basin, Iran. Climate. 6(2), 30. Farajzadeh, M., Oji, R., Cannon, A.J., Ghavidel, Y. and Bavani, A.M., 2015. An evaluation of single-site statistical downscaling techniques in terms of indices of climate extremes for the Midwest of Iran. Theoretical and Applied Climatology. 120(1-2), 377-390. Grotch, S.L. and MacCracken, M.C., 1991. The use of general circulation models to predict regional climatic change. Journal of Climate. 4(3), pp.286-303. Hewitson, B.C. and Crane, R.G., 1994. Precipitation controls in southern Mexico. In Neural Nets: Applications in Geography. 121-143. Hewitson, B.C. and Crane, R.G. eds., 1994. Neural nets: applications in geography: applications for geography. Springer Science & Business Media. (Vol. 29). Hellström, C., Chen, D., Achberger, C. and Räisänen, J., 2001. Comparison of climate change scenarios for Sweden based on statistical and dynamical downscaling of monthly precipitation. Climate Research. 19(1), 45-55. Hornik, K., 1991. Approximation capabilities of multilayer feedforward networks. Neural networks. 4(2), 251-257. Jahangir, M.H., Noorazar, L. and Azimi, E., 2019. Analyzing time series of SPI, SPEI and SPTI drought indices by using artificial neural network SOFM method and numerical comparison in chaharmahal va bakhtiari. Iranian journal of Ecohydrology. 6(3), pp.837-847. (In Persian with English abstract) Johnson, M.S., Coon, W.F., Mehta, V.K., Steenhuis, T.S., Brooks, E.S. and Boll, J., 2003. Application of two hydrologic models with different runoff mechanisms to a hillslope dominated watershed in the northeastern US: a comparison of HSPF and SMR. Journal of hydrology. 284(1-4), 57-76. Kundu, S., Khare, D. and Mondal, A., 2017. Future changes in rainfall, temperature and reference evapotranspiration in the central India by least square support vector machine. Geoscience Frontiers. 8(3), pp.583-596. Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M.I. and Huang, M., 2021. Climate change 2021: the physical science basis. Contribution of working group I to the sixth assessment report of the intergovernmental panel on climate change. p.2. Mearns, L.O., Giorgi, F., Whetton, P., Pabon, D., Hulme, M. and Lal, M., 2003. Guidelines for use of climate scenarios developed from regional climate model experiments. Data Distribution Centre of the Intergovernmental Panel on Climate Change. p.38. Mirdashtvan, M., Najafinejad, A., Malekian, A. and Sa'doddin, A.J.M.A., 2018. Downscaling the contribution to uncertainty in climate‐change assessments: representative concentration pathway (RCP) scenarios for the south alborz range, Iran. Meteorological Applications. 25(3), 414-422. Mellit, A., Pavan, A.M. and Benghanem, M., 2013. Least squares support vector machine for short-term prediction of meteorological time series. Theoretical and applied climatology. 111(1-2), pp.297-307. Murphy, J., 1999. An evaluation of statistical and dynamical techniques for downscaling local climate. Journal of Climate. 12(8), 2256-2284. Nash, JE, and Sutcliffe, JV., 1970. River flow forecasting through conceptual models part I—A discussion of principles. Journal of hydrology. 10(3), 282-90. Omidvar, E., Rezaei, M. and Pirnia, A., 2019. Performance Evaluation of Artificial Neural Network Models for Downscaling and Predicting of Climate Variables. Journal of Watershed Management Research. 9(18), pp.80-90. (In Persian with English abstract) Rezaee, M., Nahtaj, M., Moghadamniya, A. and Abkar, A., 2015. Comparison of artificial neural network and SDSM methods in the downscaling of annual rainfall in the HadCM3 modelling (case study: Kerman, Ravar and Rabor). Water Resources Engineering. 8(24), pp.25-40. (In Persian with English abstract) Sailor, D.J., Hu, T., Li, X. and Rosen, J.N., 2000. A neural network approach to local downscaling of GCM output for assessing wind power implications of climate change. Renewable Energy. 19(3), 359-378. Schoof, J.T. and Pryor, S.C., 2001. Downscaling temperature and precipitation: A comparison of regression‐based methods and artificial neural networks. International Journal of Climatology: A Journal of the Royal Meteorological Society. 21(7), 773-790. Suykens, J.A. and Vandewalle, J., 1999. Least squares support vector machine classifiers. Neural processing letters. 9(3), 293-300. Suykens, J.A., 2001, May. Nonlinear modelling and support vector machines. In IMTC 2001. Proceedings of the 18th IEEE instrumentation and measurement technology conference. Rediscovering measurement in the age of informatics (Cat. No. 01CH 37188) (Vol. 1, pp. 287-294). Suykens, J.A., De Brabanter, J., Lukas, L. and Vandewalle, J., 2002. Weighted least squares support vector machines: robustness and sparse approximation. Neurocomputing. 48(1-4), pp.85-105. Sobhani, B., Eslahi, M., Babaeian,. 2017. Comparison of statistical downscaling in climate change models to simulate climate elements in Northwest Iran. Physical Geography Research Quarterly. 49, 301-325. (In Persian with English abstract) Teegavarapu, R.S. and Goly, A., 2018. Optimal selection of predictor variables in statistical downscaling models of precipitation. Water resources management. 32(6), pp.1969-1992. Wilby, R.L., Dawson, C.W. and Barrow, E.M., 2002. SDSM—a decision support tool for the assessment of regional climate change impacts. Environmental Modelling & Software. 17(2), 145-157. Zhang, B. and Govindaraju, R.S., 2000. Prediction of watershed runoff using Bayesian concepts and modular neural networks. Water Resources Research. 36(3), 753-762.
| ||
|
آمار تعداد مشاهده مقاله: 13,066 تعداد دریافت فایل اصل مقاله: 6,575 |
||