تحلیل روند و پیش‌نگری خشکسالی در دشت مغان با بهره‌گیری از شاخص‎ spi ‎و مدل‌های اقلیمی ‎ cmip6‎‎‏

خشکسالی به‌عنوان یکی از چالش‌های مهم اقلیمی، تاثیرات گسترده‌ای بر منابع آب و کشاورزی دارد. هدف این پژوهش، پیش‌نگری وضعیت خشکسالی هواشناسی دشت مغان بر پایه تحلیل بارش در بازه‌ی 2050–2025 با استفاده از شاخص spi و مدل‌های اقلیمی نسل ششم (cmip6)، طبق دو سناریوی اقلیمی متوسط ssp2-4.5 و بدبینانه ssp5-8.5 است. در این مطالعه، داده‌‏‌های بارش مشاهده‌ای از 15 ایستگاه هواشناسی منطقه جمع‌آوری شد. سپس، 10 مدل از مدل‌های اقلیمی نسل ششم (cmip6) انتخاب گردید. با توجه به بزرگ‌مقیاس بودن خروجی مدل‌ها، فرآیند ریز‌مقیاس‌نمایی آماری و تصحیح اریبی با استفاده از ابزار cmhyd انجام شد. این فرآیند با بهره‌گیری از چهار روش مختلف تصحیح بایاس شامل مقیاس‌بندی خطی، مقیاس‌بندی شدت محلی بارش، تبدیل توانی و نگاشت توزیع، داده‌ها را به مقیاس ایستگاهی تبدیل و خطاهای سیستماتیک مدل‌ها را کاهش داد. کارایی مدل‌ها با معیارهای ارزیابی ‏r2‎، ‏mse، ‏rmse‏ و ‏mae‏ ارزیابی و از نتایج 5 مدل برتر (‏ec-earth3‎، ‏gfdl-esm4‎، ‏ec-earth3-veg، ‏miroc6‎، ‏mri-esm2-0‎‏) در محاسبه میانگین وزنی همادی این مدل‌ها برای پیش‌نگری بارش ‏دوره آینده استفاده شد و خشکسالی در مقیاس‌های 6 و 12 ماهه بررسی گردید. نتایج نشان داد که روش مقیاس‌بندی خطی بارش ‏کارآمدترین روش برای تصحیح اریبی مدل‌های ‏cmip6‎‏ بوده و مدل اقلیمی ‏ec-earth3‎‏ بهترین عملکرد را دارد. در دوره پایه، بیش‌ترین ‏بارش در ایستگاه‌های باران‌سنجی زهرا و بیله‌سوار و کم‌ترین در بران مشاهده شد. تحلیل روند شاخص ‏spi‏ با آزمون mann-kendall‏اصلاح شده نشان داد که تنها ‏ایستگاه بران در مقیاس شش‌ماهه با افزایش معنادار خشکسالی مواجه است. در افق آینده، بر اساس سناریوی ‏ssp2-4.5‎‏ خشکسالی در ‏ایستگاه‌های سینوپتیک گرمی و باران‌سنجی زهرا افزایش می‌یابد و تحت سناریوی ‏ssp5-8.5‎، ایستگاه‌های گرمی و پارس‌آباد با خشکسالی‌های شدیدتر ‏روبه‌رو خواهند شد. در مقیاس 12‌ ماهه، بهبود شاخص ‏spi‏ در بیش‌تر ایستگاه‌ها مشاهده می‌شود، اگرچه در ‏ssp5-8.5‎‏ کاهش ‏شاخص در ایستگاه‌های متعددی پیش‌بینی شده است. با این حال، طبقه نرمال خشکسالی هم‌چنان در دوره‌های آتی حاکم خواهد بود.‏

drought trend analysis and forecasting in the moghan plain using the spi index and cmip6 climate models

introductiondrought is a complex and gradual climatic phenomenon that has widespread environmental, ‎economic, and social impacts due to long-term reductions in rainfall and scarcity of water ‎resources. as global climate change intensifies, resulting in warmer and drier conditions, both the frequency and severity of ‎droughts are escalating, thereby necessitating more precise monitoring systems and comprehensive quantitative ‎analyses. indices such as spi are used to identify and assess the severity and ‎frequency of droughts. droughts are typically categorized into three distinct types, meteorological, agricultural, and hydrological, each ‎of which exerts unique impacts on natural resources and anthropogenic activities. ‎meanwhile, the moghan plain, as one of the country’s important agricultural regions, has been ‎severely affected by recent droughts. diminished precipitation, declining groundwater levels, and sparse vegetation cover collectively compromise the ‎livelihood sustainability of thousands of rural and nomadic households throughout the region.‎ materials and methodsthis study used the standard precipitation index (spi) and cmip6 climate models to predict ‎future drought in the moghan plain. precipitation data from 15 meteorological and rain gauge ‎stations were received from the iranian meteorological organization ‎(irimo.ir) and the ardabil regional water company. then, 10 valid models from the sixth cmip6 report, which are known for their superior performance in simulating precipitation parameters based on previous studies, were selected and their historical data were collected. the cmhyd downscaling model and four bias correction ‎methods were used to correct the data bias. in order to evaluate the efficiency of the models, the ‎corrected data were compared with observational data and the top five models were selected. ‎then, using the hamadi weighted average method for the top models, precipitation changes ‎were predicted for two moderate (ssp2-4.5) and pessimistic (ssp5-8.5) climate scenarios. ‎finally, the drought situation was examined at 6 and 12-month time scales and the trend of ‎changes was analyzed using the modified mann-kendall test.‎results and discussionthe performance of cmip6 models with four bias correction methods showed that the linear ‎precipitation scaling method has the best results. the ec-earth3 model was recognized as the ‎superior model, followed by the gfdl-esm4, ec-earth3-veg, miroc6, and mri-esm2-0 ‎models. spatio-temporal analysis of annual precipitation in the moghan plain shows a decrease in ‎the precipitation pattern from the southeast to the northwest. projections indicate an increase in annual precipitation and a decrease in precipitation in warm months under ssp2-4.5 (medium emission) and ssp5-8.5 (high emission) scenarios, which could jeopardize the sustainability of agriculture in the moghan plain by affecting water availability during the growing season. analysis of the six-month spi ‎index during the observation period shows an increasing trend and a decrease in drought ‎intensity in most stations. however, in the ssp2-4.5 scenario, garmi and zahra stations and in ‎ssp5-8.5, garmi, dasht and parsabad stations experience a decrease in spi and an increase in ‎drought. the results of the 12-month spi analysis also confirm an increasing trend in most ‎stations and a decrease in drought intensity during the observation period. in ssp2-4.5, no significant changes are observed, indicating a relatively stable climate. however, under ssp5-8.5, some stations such as agha mohammad biglo and dasht show a significant decreasing trend, pointing to more severe and prolonged droughts. the analysis of the frequency of the six-month spi index (period ‎‎1985–2014) shows that the normal condition was the most frequent, followed by drought and ‎mild and moderate wetness, while severe events were observed less frequently. in the period 2025–2050, under ssp2-4.5, the frequency of drought and mild wetness events is expected to increase, while severe events are projected to decrease. in ssp5-8.5, despite the increase in precipitation, the ‎concentration of mild and moderate droughts will increase in autumn. the climatic sequence ‎during the observation period shows that the normal class has the most stable condition and ‎droughts are mainly short-term. in the future, long-term normal sequences will continue, but ‎some areas of the moghan plain will face a higher risk of drought. in the statistical and spatial ‎analysis of the observation period with a 12-month scale, the normal class has the highest ‎frequency, indicating the relative stability of the climate. the stations of qarakhan biglo and ‎parsabad are known as unstable centers, and the zahra station is known as highly stable. in the ‎ssp2-4.5 scenario, the normal class continues to dominate, but seasonal fluctuations increase, ‎especially in summer. stations such as oslandoz and shourestan will have more severe ‎fluctuations. in ssp5-8.5, similar conditions are also seen with greater intensity. the months of ‎june and september are relatively more stable, while march and august show the peak of ‎drought and wetness. the drought classification time series shows that the normal class tends to persist for longer periods, especially in five-month and long-term durations. meanwhile, short-term mild and moderate droughts are frequently observed in most stations. in the future, although the pattern of frequency ‎and sequence will be relatively stable, the severity of extreme conditions will increase at some ‎stations.‎conclusion in this study, the linear precipitation scaling method was identified as the most effective ‎technique for bias correction of cmip6 climate models in simulating the rainfall of the moghan ‎plain. the ec-earth3 model showed superior performance and was used together with four ‎other models to build the hamadi model. the results of the base period (1985–2014) confirmed ‎the spatial pattern of precipitation with maximum in the east and southeast and minimum in the ‎west of the region. climate projections for the period 2025–2050 indicate an increase in annual precipitation in most areas, especially in the west of the region, while summer precipitation is expected to decrease. the standardized drought index (spi) at six-month and twelve-month scales showed ‎that in the ssp2-4.5 scenario, relative climate stability is maintained, but in ssp5-8.5, the ‎probability of more severe droughts increases. despite the continued dominance of the normal ‎class, seasonal fluctuations and the intensity of extreme phenomena will increase at some ‎stations.‎