شبیه‌سازی و پیش‌بینی برخی متغیرهای اقلیمی توسط مدل چندگانه خطی sdsm و سناریوهای rcp در حوضه آبخیز حاجیلر

مدل‌های ریزمقیاس نمایی به‌منظور تولید مقادیر بارش واقعی در یک محدوده محیطی از طریق شبیه‌سازی مجموعه داده‌های اتمسفری مورداستفاده قرار می‌گیرند. هدف از این پژوهش پیش‌بینی مقادیر عناصر اقلیمی دما و بارش ایستگاه سینوپتیک اهر با مدل ریزمقیاس نمایی sdsm و با استفاده از خروجی مدل تغییر اقلیمی canesm2 تحت سه سناریو 2.6 rcp، 4.5 rcp و 8.5 rcp ریزمقیاس و برای چهار دوره 18 ساله برای دوره‌های آینده 2020-2040،2039-2060،2059-2079 و 2080- 2099 بوده است. همچنین از شاخص های آماری میانگین جذر خطا مربع (rmse) و میانگین خطای مطلق و (mae) استفاده شده و بررسی روند سالانه این تغییرات با استفاده از آزمون کولموگروف اسمیرنوف و شاپیرو-ویلک است. نتایج پیش‌بینی ها نشان داد در سناریوهای 2.6 rcp و 4.5 در فصل بهار کاهش بارش به میزان 14 تا 5 میلی‌متر در ایستگاه سینوپتیک اهر رخ می‌دهد همچنین متوسط دمای حداکثر از 0.1 تا 1.56 درجه سانتی گراد، متوسط دمای حداقل از 0.4 تا 1.54 درجه سانتی گراد در هر سه سناریو در خلال دوره‌های زمانی افزایش خواهد یافت. از نظر روند تغییرات بارش فصلی در تمام سناریوها در فصول بارشی کاهش و در ماه‌های سرد افزایش داشته و در مورد عنصرهای دما در تمام سال این روند معنادار و افزایشی خواهد بود.

simulation and forecasting of some climatic variables by sdsm multiple linear model and rcp scenarios in hajiler watershed

introduction climate changes have occurred throughout earth’s history. however, in recent years, human interference in the environment has accelerated temperature changes, leading to more pronounced climate shifts. according to the fourth assessment report of the intergovernmental panel on climate change (ipcc), the average global temperature increased by ±0.18 to 0.7°c over the past century (1906 to 2007) (ipcc, 2007). furthermore, the fifth assessment report (ipcc, 2013) indicates an increase of 0.85°c between 1901 and 2012. global warming, with its numerous negative impacts on biological systems, is a critical issue today. to assess these changes at a regional level, detailed investigations using downscaling techniques are necessary. among these, statistical downscaling models (e.g., sdsm) have proven effective for simulating and evaluating climate change impacts. despite extensive research on downscaling, no prior study has specifically examined the exponential downscaling of maximum and minimum temperature and precipitation for the hajiler watershed using ipcc’s fifth report scenarios. therefore, this study aims to simulate monthly temperature and precipitation parameters at ahar synoptic station using the sdsm model under rcp scenarios and analyze their annual trends using the non-parametric mann-kendall test.material and methods this study employed the sdsm model (version 5.3) and three rcp scenarios (2.6, 4.5, and 8.5) from the ipcc fifth assessment report. climate data from ahar synoptic station were used to simulate precipitation and temperature (maximum and minimum) across four time periods: the near future (2020-2039), mid-future (2040-2059, 2060-2079), and far future (2080-2099). in the first stage, daily data for precipitation and temperature from 1986 to 2005 were collected from the meteorological organization and underwent quality control. these data were then used in the sdsm model for monthly-scale simulations and compared with baseline observations.the sdsm model incorporates three types of data for downscaling: daily observational data (predictand), ncep reanalysis data (predictor), and large-scale forecast data from atmospheric general circulation models (gcms). predictor data from the canesm2 model were sourced from the environment canada website. large-scale daily time series data (1961-2005) for precipitation, maximum temperature, and minimum temperature were pre-processed in excel and prepared for analysis in notepad. predictor variables suitable for the region were selected based on correlation with observational data and analyzed using the ncep and canesm2 datasets. using these predictors, precipitation and temperature values for future scenarios (rcp 2.6, rcp 4.5, and rcp 8.5) were simulated for the four study periods.results and discussion to predict temperature changes in the hajiler watershed, the sdsm statistical downscaling model was used. the statistical relationship between observed and predictor variables was evaluated based on correlation coefficients. among the 26 atmospheric variables tested, nceptempgl and ncepp500gl exhibited the highest correlation with temperature and precipitation data. temperature data showed a stronger correlation with observational data than precipitation due to temperature’s continuous nature and lower variability compared to precipitation, which is more influenced by anomalies.after validating the model for the baseline period (1986-2005), climatic parameters were simulated for future periods under the rcp 2.6, rcp 4.5, and rcp 8.5 scenarios using the canesm2 global model. the analysis revealed that climate change affects precipitation in two primary ways: changes in precipitation amount and changes in its temporal distribution. results indicate that precipitation patterns will become increasingly irregular, with rainfall occurring during inappropriate seasons.