تحلیل عدم قطعیت مدل‌های تغییر اقلیم در پیش‌بینی دمای متوسط ماهانه با استفاده از ابرمکعب لاتین (مطالعه موردی: حوزه آبخیز سد میناب)

ارزیابی اثر تغییر اقلیم در دهه‌های آینده با هدف برنامه‌ریزی محیطی و کاهش اثرات آن امری ضروری است. در مطالعات تغییر اقلیم، لحاظ نکردن عدم قطعیت‌های موجود در مراحل مختلف ارزیابی اثرات، سبب کاهش قطعیت اطمینان به خروجی‌های نهایی سیستم خواهد شد. این عدم قطعیت ناشی از کارکرد مدل‌های گردش عمومی، سناریوهای مختلف انتشار و فرآیند ریزمقیاس نمایی است. در این پژوهش، عدم قطعیت تغییرات دمای متوسط ماهانه حوزه آبخیز سد میناب در دو دوره زمانی (2045-2016 و 2075-2046) و بر اساس خروجی پنج مدل اقلیمی (hadgem2es، bnuesm، ccsm4، csiromk36 و mpiesmmr) و سه سناریو rcp2.6، rcp4.5 و rcp8.5 مورد بررسی قرار گرفت. بدین منظور با استفاده از روش change factor متغیر دمای متوسط برای دوره‌های آینده، مقیاس‌کاهی گردید. جهت بررسی عدم قطعیت مدل‌ها در سه سناریو و دوره مورد نظر، از روش ابر مکعب لاتین که یک روش نمونه‌برداری تصادفی طبقه‌بندی است، استفاده گردید. در برررسی عدم قطعیت دوره‌ها، در تمامی مدل‌‌ها و سناریوها، عدم قطعیت دوره دوم (2046-2075) در برآورد دما بیشتر از دوره اول (2016-2045) است. بدین مفهوم که افزایش طول دوره نسبت به دوره مشاهداتی سبب افزایش خطا در پیش‌بینی مدل‌های تغییر اقلیم می‌گردد. در بررسی مدل‌ها نیز، کمترین عدم قطعیت مربوط به مدل csiromk36 در سناریوی rcp2.6 و دوره 2016-2045 و بیشترین عدم قطعیت مربوط به مدل hadgem2es در سناریوو دوره مذکور است.

Uncertainty analysis of global climate models in predicting monthly average temperature using Latin Hypercub Sampling (case study: Minab Dam basin)

It is necessary to understand the climate change in next decades to have a suitable environmental planning for adapting and reducing it’s effects. In climate change studies, ignoring uncertainties at various stages of impact assessment will reduce confidence in system results. This uncertainty is due to the performance of general circulation models, different emission scenarios and doanscaling process. In this research uncertainty of monthly average temperature of drainage basin of Minab dam is projected in two periods(20162045 and 20462075) using outcomes of the five general circulation models of the HADGEM2ES, BNUESM, CCSM4, CSIROMK36, MPIESMMR under three scenarios of RCP2.6, RCP4.5 and RCP8.5. For this purpose, using the method of variants of average tempreture for change factor, future is downscaled. Latin hypercube method wich is an accidental sampelling is used here for checking the uncertainty of models. The results of period uncertainty, in all models and scenarios showed that, the uncertainty of the second period (20462075) is greater than in the first period (20162045). It means, increasing the length of the forecast period increases the error in predicting climate change models. The results showed the uncertainty of the different models showed that the least uncertainty was related to the CSIROMk36 model in the RCP2.6 scenario and the 20162045 period, while the highest uncertainty was related to the HadGEM2ES model in the scenario and period.It is necessary to understand the climate change in next decades to have a suitable environmental planning for adapting and reducing it’s effects. In climate change studies, ignoring uncertainties at various stages of impact assessment will reduce confidence in system results. This uncertainty is due to the performance of general circulation models, different emission scenarios and doanscaling process. In this research uncertainty of monthly average temperature of drainage basin of Minab dam is projected in two periods(20162045 and 20462075) using outcomes of the five general circulation models of the HADGEM2ES, BNUESM, CCSM4, CSIROMK36, MPIESMMR under three scenarios of RCP2.6, RCP4.5 and RCP8.5. For this purpose, using the method of variants of average tempreture for change factor, future is downscaled. Latin hypercube method wich is an accidental sampelling is used here for checking the uncertainty of models. The results of period uncertainty, in all models and scenarios showed that, the uncertainty of the second period (20462075) is greater than in the first period (20162045). It means, increasing the length of the forecast period increases the error in predicting climate change models. The results showed the uncertainty of the different models showed that the least uncertainty was related to the CSIROMk36 model in the RCP2.6 scenario and the 20162045 period, while the highest uncertainty was related to the HadGEM2ES model in the scenario and period.It is necessary to understand the climate change in next decades to have a suitable environmental planning for adapting and reducing it’s effects. In climate change studies, ignoring uncertainties at various stages of impact assessment will reduce confidence in system results. This uncertainty is due to the performance of general circulation models, different emission scenarios and doanscaling process. In this research uncertainty of monthly average temperature of drainage basin of Minab dam is projected in two periods(20162045 and 20462075) using outcomes of the five general circulation models of the HADGEM2ES, BNUESM, CCSM4, CSIROMK36, MPIESMMR under three scenarios of RCP2.6, RCP4.5 and RCP8.5. For this purpose, using the method of variants of average tempreture for change factor, future is downscaled. Latin hypercube method wich is an accidental sampelling is used here for checking the uncertainty of models. The results of period uncertainty, in all models and scenarios showed that, the uncertainty of the second period (20462075) is greater than in the first period (20162045). It means, increasing the length of the forecast period increases the error in predicting climate change models. The results showed the uncertainty of the different models showed that the least uncertainty was related to the CSIROMk36 model in the RCP2.6 scenario and the 20162045 period, while the highest uncertainty was related to the HadGEM2ES model in the scenario and period.It is necessary to understand the climate change in next decades to have a suitable environmental planning for adapting and reducing it’s effects. In climate change studies, ignoring uncertainties at various stages of impact assessment will reduce confidence in system results. This uncertainty is due to the performance of general circulation models, different emission scenarios and doanscaling process. In this research uncertainty of monthly average temperature of drainage basin of Minab dam is projected in two periods(20162045 and 20462075) using outcomes of the five general circulation models of the HADGEM2ES, BNUESM, CCSM4, CSIROMK36, MPIESMMR under three scenarios of RCP2.6, RCP4.5 and RCP8.5. For this purpose, using the method of variants of average tempreture for change factor, future is downscaled. Latin hypercube method wich is an accidental sampelling is used here for checking the uncertainty of models. The results of period uncertainty, in all models and scenarios showed that, the uncertainty of the second period (20462075) is greater than in the first period (20162045). It means, increasing the length of the forecast period increases the error in predicting climate change models. The results showed the uncertainty of the different models showed that the least uncertainty was related to the CSIROMk36 model in the RCP2.6 scenario and the 20162045 period, while the highest uncertainty was related to the HadGEM2ES model in the scenario and period.