پیش‌نگری شدت بارش در ایران با به‌کارگیری رویکرد همادی چندمدلی با استفاده از داده‌های مقیاس‎‌کاهی‌شده nex-gddp

هدف از این مطالعه بررسی کارایی مدل های مقیاس‌کاهی شده روزانه جهانی تبادل زمین ناسا  (nex - gddp) در شبیه سازی شدت بارش و پیش‌نگری بلندمدت آن در ایران است. برای این منظور نُه مدل از مدل های cmip5 از پروژه nex - gddp بر اساس حساسیت اقلیمی گزینش شد. برای درستی سنجی برونداد بارش داده های مذکور از داده های بارش 49 ایستگاه همدیدی طی دوره تاریخی (2005 - 1980) و دو سنجه آماری rmse و mbe استفاده شد. نتایج نشان داد داده های پروژه nex - gddp در مقایسه با داده های مشاهداتی اریبی چندانی ندارند و بیشتر مدل ها با خطای نسبی کم، کارایی لازم را در بازتولید الگوی فضایی بارش در ایران دارند. از بین مدل های نُه‌گانه بررسی شده، مدل mpi - esm - lr بیشینه بیش برآوردی و مدل ipsl - cm5 - alr بیشینه کم‌برآوردی را در ایران نشان می دهد. در مقایسه با سایر gcm ها در دوره تاریخی، داده‌های پروژه nex - gddp عدم قطعیت کمتری را در مقیاس منطقه‌ای نشان می دهند و از این‌رو پیش‌نگری های nex - gddp بسیار مطمئن تر است. از روش میانگین گیری مدل بیزی (bma) جهت تولید یک مدل همادی از مدل های نُه‌گانه استفاده شد. بر اساس مساحت زیر خم roc، مدل همادی تولید شده کارایی بهتری را نسبت به مدل های منفرد نشان داد. پیش نگری شدت بارش با دو شاخص sdii و rx1day با مدل همادی nex - gddp - mme طی سه دوره آینده نزدیک (2050 - 2026)، آینده میانی (2075 - 2051) و آینده دور (2100 - 2076) با دو سناریوی rcp4.5 و rcp8.5 انجام شد. پیش‌نگری های شدت بارش نشان می دهد در آینده در سراسر ایران بارش با شدت بیشتری اتفاق می افتد. شاخص‌های rx1day و sdii تا پایان قرن در حدود 4 تا 13 درصد برای متوسط پهنه ایران افزایش خواهند یافت که نشان دهنده افزایش بارش‌های سیل‌آسا طی دهه های آینده در ایران است.

projection of precipitation intensity in iran using nex-gddp by multi-model ensemble approach

global warming has a significant impact on weather and climate change. these changes, and especially changes in climate extremes, have a great impact on human society and ecosystems. future changes in extreme climate events, including precipitation extreme, will cause great damage to society, the economy, and ecosystems because of their potentially severe effects. the purpose of this study is to investigate the performance of nasa earth exchange global daily downscaled projections (nexgddp) in simulating precipitation and its longterm projection in iran.    for this purpose, the nine models of nexgddp were selected based on climate sensitivity. precipitations from 49 ground stations during the historical period (19802005) were used to evaluate the precipitation output of the mentioned models using rmse and mbe statistics. the bayesian model averaging (bma) method was used to generate an ensemble model from nine models. intensity of precipitation with two indices sdii and rx1day is projected during the three periods of near future (20262050), medium future (20512075) and far future (20762100) under two scenarios rcp4.5 and rcp8.5.    the results showed that nexgddp models did not have much bias compared to observation and most models with low relative error have good performance in reproducing the spatial pattern of precipitation in iran. among the nine selected models, mpiesmlr model has shown the maximum overestimation and ipslcm5alr model has shown the maximum underestimation in iran. compared to other gcms in the historical period, nexgddp models show less uncertainty at the regional scale;therefore, nexgddp simulations are much more reliable. the precipitation intensity projections show that in the future, precipitation will occur more intensively throughout iran. the rx1day and sdii indices will increase by about 4 to 13 percent for the average area of iran by the end of the century, which indicates an increase in flooding in iran in the coming decades.    projections of precipitation intensity in iran based on two indices rx1day and sdii from the set of precipitation index of etccdi working group by ensemble model nexgddpmme showed that with the continuation of global warming, precipitation intensity will increase significantly throughout iran. the maximum oneday precipitation amount (rx1day) will increase between 4.42 to 13.08 percent for the areaaveraged by the end of this century compared to 19802005. moreover, the sdii index will increase between 4.45 to 13.96% for areaaverage of iran. the highest increase in precipitation intensity generally occurs in the coastal region of southern iran, especially in the coasts of the persian gulf and western iran, while the lowest increase is generally observed in the northwestern region. introductionglobal warming has a significant impact on weather and climate change. these changes, and especially changes in climate extremes, have a great impact on human society and ecosystems. future changes in extreme climate events, including precipitation extreme, will cause great damage to society, the economy, and ecosystems because of their potentially severe effects. the purpose of this study is to investigate the performance of nasa earth exchange global daily downscaled projections (nexgddp) in simulating precipitation and its longterm projection in iran.materials and methodsfor this purpose, the nine models of nexgddp were selected based on climate sensitivity. precipitation from 49 ground stations during the historical period (19802005) were used to evaluate the precipitation output of the mentioned models using rmse and mbe statistics. the bayesian model averaging (bma) method was used to generate an ensemble model from nine models. intensity of precipitation with two indices sdii and rx1day is projected during the three periods of near future (20262050), medium future (20512075) and far future (20762100) under two scenarios rcp4.5 and rcp8.5.results and discussionthe results showed that nexgddp models did not have much bias compared to observation and most models with low relative error have good performance in reproducing the spatial pattern of precipitation in iran. among the nine selected models, mpiesmlr model has shown the maximum overestimation and ipslcm5alr model has shown the maximum underestimation in iran. compared to other gcms in the historical period, nexgddp models show less uncertainty at the regional scale, and therefore nexgddp simulation are much more reliable. the precipitation intensity projections show that in the future, precipitation will occur more intensively throughout iran. the rx1day and sdii indices will increase by about 4 to 13 percent for the average area of iran by the end of the century, which indicates an increase in flooding in iran in the coming decades.conclusionprojections of precipitation intensity in iran based on two indices rx1day and sdii from the set of precipitation index of etccdi working group by ensemble model nexgddpmme showed that with the continuation of global warming, precipitation intensity will increase significantly throughout iran. the maximum oneday precipitation amount (rx1day) will increase between 4.42 to 13.08 percent for the areaaveraged by the end of this century compared to 19802005. also, the sdii index will increase between 4.45 to 13.96% for areaaverage of iran. the highest increase in precipitation intensity generally occurs in the coastal region of southern iran, especially in the coasts of the persian gulf and western iran, while the lowest increase is generally observed in the northwestern region.keywords