Projection of the future outlook of temperature and precipitation in Urmia Lake basin by the CMIP6 models

ABSTRACT
In this research, the climatic variables of temperature and precipitation of Urmia Lake catchment area were evaluated and projected using CMIP6 models under two scenarios SSP1-2.6 and SSP5-8.5 in the future periods (2031-2055 and 2071-2095). First, the accuracy of the models for the base period (1990-2014) was evaluated using Taylor diagram and RMSE and NRMSE indices after downscaling with different methods of quantile mapping, and among the models, the MRI-ESM2-0 for temperature and INM-CM5-0 for precipitation using the SSPLIN downscaling method were selected to project the future climate, then the future temperature and precipitation data were produced. The results of comparing the temperature and precipitation of the future periods with the base period showed that the average annual temperature of the basin will increase under all scenarios. The average annual temperature increase of the basin in the near future in the optimistic and pessimistic scenarios will be 1.5 and 1.8 degrees Celsius, respectively, and in the far future, 1.4 and 4 degrees Celsius respectively. The average annual precipitation will decrease in all scenarios, in the near future it will decrease by 19.9 and 21.6 percent in the optimistic and pessimistic scenarios, respectively, and by 12 and 28.6 percent in the far future, respectively. Based on the spatial distribution of changes in temperature and precipitation in the future periods, the greatest increase in temperature and the greatest decrease in precipitation will occur in the northern areas of the basin.
Extended Abstract
Introduction
Currently, the World Climate Research Program (WCRP) has launched the sixth phase of CMIP models. These models have higher spatial resolution and improved parameterization schemes for the main physical and biogeochemical processes of the climate system. To project future climate, CMIP6 models use emission scenarios based on common socio-economic pathways (SSP). These scenarios are developed based on the emission and land use routes, which are somewhat different from the RCP scenarios. The advantage of SSP scenarios is that they have a clear description of the socio-economic evolution of the future society (Zhou et al., 2021). The sixth assessment report of the IPCC states that global warming will continue and addressing the challenges caused by human-induced climate change has become the main issue of the 21st century (IPCC, 2021). Due to environmental and geographical factors in different regions of the world, these regions will face different challenges. Therefore, regional climate change studies are vital (Liu et al., 2022).
The outputs of atmospheric general circulation models in terms of temporal and spatial resolution are about tens of kilometers on a daily and monthly scale, which are large scale compared to climatic and hydrological processes. In addition, GCM simulations in both spatial and temporal scales have uncertainty in the parameterization of processes, so the output of these models cannot be directly used in climate change studies. Therefore, downscaling and bias correcting of GCM simulations is necessary to obtain information at the appropriate scale (Wood et al., 2004). The reason for using models with a higher ranking in terms of simulation skill is that the models have different skills in different regions and periods (Bağçaci et al., 2021) and using the average of several different models are effective in reducing the uncertainty of the results of Simulations.
Several studies have been conducted in the field of climate change forecasting and GCM models. Based on these studies, it can be seen that the changes in climatic parameters in different regions have different characteristics and effects, and the reason for that is the different climatic and geographical characteristics of those regions. On the other hand, the use of different downscaling models and methods has been effective in the obtained results. Therefore, in studying and forecasting the climate of each region, the use of new models, validation of models, selection of the most appropriate models, and the use of appropriate downscaling methods increase the validity of the research and its results It can be used in different fields.
 
Methodology
Urmia Lake basin with an area of about 52,000 km² is located in the northwest of Iran and includes parts of East Azerbaijan, West Azerbaijan and Kurdistan provinces.
In order to project the future temperature and precipitation using CMIP6 models, first the accuracy and performance of the models for simulating the base period in comparison with the observational data were evaluated by using downscaling statistical methods.
For this purpose, the output of GCM models for the base period was divided into two periods: 1990-2004 for calibration and 2005-2014 for validation. In the calibration phase, the statistical correlation between the observational data and the model were determined, and assuming that the resulting relationships were established for the future period, the data for the future period (2005-2014) were generated. In the validation phase, the accuracy of the data produced by the downscaling model compared to the observational data was evaluated using the graphs of the monthly average values of the variables, the Taylor diagram, and the RMSE and NRMSE indices. After evaluating the accuracy of the models, the model with the least error in simulating temperature and precipitation, was used to project the future climate.
 
 
 
Results and Discussion
The average monthly temperature of the basin will increase in most months of the year under all scenarios; the highest increase in temperature will occur in the months of September and October and the lowest increase in spring. Rainfall changes will be different in different months. In April and May (spring season) and October and November (autumn season) in all scenarios, the rainfall of the basin will decrease. The highest percentage of precipitation reduction will be in the autumn season.
The average annual temperature increase of the basin in the near future will be 1.5 and 1.4 T⁰C in the optimistic and pessimistic scenarios, and 1.8 and 4 T⁰C in the far future. Therefore, the temperature increase in the pessimistic scenario is more than the optimistic scenario. The amount of annual precipitation will decrease in all scenarios, in the near future in the optimistic and pessimistic scenarios by 19.9 and 21.6 percent respectively, and in the far future in the optimistic and pessimistic scenarios will decrease by 12 and 28.6 percent respectively.
 
Conclusion
According to the results of the present research, in the catchment area of Urmia Lake, the temperature will generally increase and precipitation will decrease in the future periods. The results of the increase in temperature and decrease in rainfall are in line with most of the previous studies in different regions. Some previous studies have been conducted in relation to the projecting of temperature and precipitation in the catchment area of Urmia Lake. Karimi and Nabizadeh (2017) using the HadCM3 model have projected an increase in temperature and a decrease in precipitation, and Zarin and Dadashi (2019) have projected an increase in temperature for the catchment area of Urmia Lake using CMIP6 models, and the results of the aforementioned studies are in accordance with the results of the present research and confirm each other's results.
 
Funding
There is no funding support.
 
Authors’ Contribution
All of the authors approved thecontent of the manuscript and agreed on all aspects of the work.
 
Conflict of Interest
Authors declared no conflict of interest.
 
Acknowledgments
We are grateful to all the scientific consultants of this paper.