پیش بینی تغییرات میانگین دمای فصلی شهرستان الشتر با استفاده ازمدل شبکه عصبی ومدل سری زمانی آریما

پیش بینی در د نیای کنو نی جز لاینفک زندگی بشر محسوب می گردد .پیش بینی روند دما نسبت به سایر پارامترهای اقلیمی در مطالعات محیطی و جوی از اهمیت ویژ ه ای برخوردار می باشد، زیرا در صنعت ، خشکسالی ، تبخیر وتعرق کار برد و فراوانی دارد . هدف ازاین پزوهش، پیش نگری نوسانات دما در فصل های سرد سال برای یازده سال آینده (2029-2019) با استفاده از مدل شبکه عصبی مصنوعی و سری زمانی آریما ( auto arima)و مقایسه مدل های نامبرده در شهرستان الشتر واقع در استان لرستان است . برای تحقق هدف فوق ؛ آمار اقلیمی 12 ایستگا ه سینوپتیک در استان لرستان مورد مطالعه قرار گرفت . داده های اقلیمی دما در یک دوره آماری30ساله از سال( 2010- 1980) از سازمان هواشناسی کشورتهیه شد . پارامترهای مورد استفاده در مدل های فوق شامل میانگین حداقل وحداکثر دمای فصلی می باشند . که با استفاده از مدل سازی شبکه عصبی مصنوعی ، سری زمانی آریما از طریق لایه های ورودی ، مخفی ، خروجی به وسیله نرون وپرسپترون ، به پیش نگری تغییرات میانگین دمای فصلی می پردازند . محاسبات میانگین تغییرات دمای فصلی در بازه زمانی (2018 1998 با پکیچ ( forecasts فرکست و شاخص rmse تحلیل گرnnar انجام شد . نمودارها و گراف ها ترسیم شده است و نتایج بدست آمده جهت پیش نگری دمای فصلی در مقایسه مدل قید شده با دقت 95 80 درصدی نشان دهنده آنست که بیشترین دقت اندازه گیری پیش نگری دما در فصل تابستان با 33% وکمترین دقت اندازه گیری در فصل پاییز با 81% می باشد . نشان از مقایسه دو مدل ذکر شده مشخص شد که مدل شبکه عصبی کارایی بهتر ی نسبت به مدل آریما بر خوردار است .

a case study of predicting seasonal average temperature changes in al shatar city using neural network model and arima time series model

the most important pillar of a scientific and applied research is the statement of the problem, when a problem can be scientific and practical that creates a challenge in relation to the solution of the problem and clearly defines the purpose of the work, as well as the challenges that have arisen in relation to the problem in question, the researcher uses simulation models can overcome one of the challenges and work as a source of information for climate researchers to use for future research. in this research, the statement of the present problem is the statement of forecasting the average seasonal temperature. what element is temperature, the answer to these questions and reasons, let’s hypothesize against it, after formulating the assumptions, prepare climatic data of temperature of the study area and the neighboring stations of the area, and also specify the study area to start the work, using modeling (simulation) and comparison and accuracy of forecasting, he used two models by comparing and measuring the accuracy of their errors, because temperature is a physical quantity, some of the sun’s radiant energy is absorbed by the earth’s surface and becomes thermal energy.this energy is expressed in the form of temperature or degrees. among the different climatic elements, temperature and precipitation are of special importance to predict this. the important key climatic element, our goal is to examine the seasonal average temperature changes in the seasons and determine the seasonal changes with 95% and 80% accuracy using artificial neural network arima time series model, rmse index, and also the models together let’s compare which predicts temperature changes better. so that researchers can use and test these models in future researches to predict other climate parameters and also the impactful consequences of seasonal temperature changes and climate elements such as relative humidity evaporation and transpiration industry transportation bridges and other infrastructures. proper planning and management should be done in this regard. in the 21st century, climate change is considered one of the biggest environmental threats to the world. changes in farin’s climate are estimated to have more negative effects on human society and the natural environment than changes in the average climate (mahmood and babel, 2014: 56). based on the fourth report of the international commission on climate change, which was published under the title of climate change assessment reports, the global increase in temperature and the occurrence of climate change have been confirmed by using the measured data of the surface temperature of land and water in the world (ipc si, 2014: 32). the first effect of climate change on atmospheric elements is especially temperature and precipitation, then due to the relationship between atmospheric elements and terrestrial ecosystems, water resources, vegetation, soil and also human life will be affected by this phenomenon; therefore, investigating the trend of atmospheric variables such as temperature is of particular importance (abkar et al., 2013: 14).temperature some of the radiant energy of the sun absorbed by the effects of the earth’s surface turns into thermal energy. this energy is manifested in the form of temperature or degree. among the different climatic elements, temperature and precipitation are of particular importance.