پیش‌بینی دمای ماهانه ایران با استفاده از پروژه پیش‌بینی اقلیمی دهه‌ای (dcpp) در دهه‌ آینده (2028-2021)

این پژوهش با هدف درستی‌‌سنجی و پیش‌بینی دهه ای دمای ماهانه ایران با استفاده از برونداد پروژه پیش‌بینی اقلیمی دهه ای (dcpp) که در پروژه مقایسه مدل های جفت‌شده فاز ششم (cmip6) مشارکت داده شده است، انجام شد. به این منظور از دو گروه داده شامل دمای 42 ایستگاه همدید و برونداد دو مدل bcc-csm2-mr و mpi-esm1-2-hr با تفکیک افقی 100 کیلومتر برای دوره گذشته‌نگر (2019-1980) و دوره پیش بینی (2028-2021) استفاده شد. برای درستی‌‌سنجی مدل ها از سنجه‌های آماری rmse، mbe و pcc و نمودار تیلور استفاده شد. نتایج درستی‌‌سنجی نشان‌دهنده کارایی بهتر مدل mpi-esm1-2-hr نسبت به مدل bcc-csm2-mr در ایران می باشد. در مقابل مدل bcc-csm2-mr به‌خصوص در سواحل شمالی و جنوبی کشور با خطای قابل‌توجهی همراه بوده و صرفاً در مناطق کوهستانی ایران عملکرد قابل‌قبولی را نشان داده است. نتایج بررسی دمای ماهانه دوره گذشته نگر با استفاده از مدل mpi-esm1-2-hr نشان داد که این مدل به‌خوبی الگوهای دمایی کشور را به تصویر می کشد. پیش بینی دمای ایران برای دوره 2028-2021 نشان داد که بی هنجاری دما در دهه ی آینده در تمامی ماه های سال مثبت و به‌طور متوسط 0.9 درجه سلسیوس افزایش می یابد. آنچه در این میان بسیار حائز اهمیت است بی هنجاری مثبت دمای ماه های مارس، آوریل، ژوئن و ژوئیه در تمامی پهنه های اقلیمی کشور است. بی هنجاری مثبت در این ماه ها بیش از یک درجه سلسیوس بوده که یک تهدید جدی برای محیط‌زیست و منابع آب ایران در سال های آینده به شمار می رود.

Near term (2021-2028) climate prediction of monthly temperature in Iran using Decadal Climate Prediction Project (DCPP)

Decadal prediction is a general term that encompasses predictions for annual, interannual, and decadal periods in which significant progress has been made over the years. Decadal climate prediction is made using a hindcast and the latest generation of climate models. It provides two categories of hindcast and prediction data. The purpose of this study is to evaluate the temperature from the DCPP and its prediction in Iran based on the available models of the DCPP project contribution to the CMIP6 project.The study area of this research is Iran. As mentioned, the purpose of this study is to predict the nearterm temperature based on the output of the DCPP project. For this purpose, daily temperature from 42 synoptic stations was used as observation to evaluate the available models of the DCPP project. Unlike general circulation models (GCMs), the DCPP project has an initialization that includes a threemonth time step for implementation of each year. Air temperature of two models BCCCSM2MR and MPIESM12HR with a horizontal resolution of 100 km is available for the DCPP project from the CMIP6 series. Three statistics, Pearson correlation coefficient (PCC), root mean square error (RMSE) and mean bias error (MBE), were used to evaluate the selected models of the DCPP project using observational data (synoptic stations).In the study of the relationship between observation and hindcast of the two selected models, it is found that the BCCCSM2MR model shows a high correlation (0.99) in the mountainous areas of Zagros and Alborz and arid and semiarid regions of the inland and east of Iran. However, the northern and southern coasts show a weak correlation (between 0.92 and 0.97). Examination of RMSE statistics for the BCCCSM2MR model also shows the maximum error between 1.2 to 2.2o in the coastal areas of the country (the Caspian Sea and the Oman Sea). The western and northern mountains of Iran show the minimum RMSE.The BCCCSM2MR model shows more bias than the MPIESM12HR model in the northern regions of the country. Examination of the average monthly temperature anomaly across Iran in the predicted period compared to the hindcast period (19802019) showed that the monthly temperature anomaly is positive across the country compared to the normal period in all months of the year. This value is 1.03 degrees Celsius for the countrywide average. In other words, the temperature in Iran will increase by one degree for the bear term period (20212028) compared to the longterm period of the last 40 years (19802019).In this study, for the first time, a decadal climate prediction of Iran’s monthly temperature is assessed using the output of two available models BCCCSM2MR and MPIESM12HR from the DCPP contribution to the Coupled Model Intercomparison Project Phase 6 (CMIP6). The evaluation of the models using three statistical measures RMSE, MBE and PCC showed that the BCCCSM2MR model has the lowest performance in the coastal areas of Iran (the Caspian and the Oman Sea) and the highest performance in the highlands of Iran. The output of the MPIESM12HR model during the hindcast period (19802019) show good performance of this model in determining the temperature patterns of the country. The minimum temperature is based on the output of this model in January with a value of 6.28o. Examination of the predicted temperature anomaly (20212028) compared to the hindcast period (19802019) shows that the average anomaly across the country for different months of the year during 20212028 compared to the hindcast period is 0.99o.