Projection the Wind Field in the Future Based on the CMIP5 and CMIP6 Climate Models in Sistan and Baluchestan Province Introduction

Introduction

Renewable energy plays a crucial role in reducing greenhouse gas emissions and combating global climate change. Among the various renewable energy sources, wind energy stands out due to its production capacity and rapid technological advancement. This study aims to evaluate the capabilities and uncertainties of CMIP5 and CMIP6 models under two scenarios—RCP4.5 and RCP8.5 for CMIP5, and SSP2-4.5 and SSP5-8.5 for CMIP6—in simulating wind speed. Additionally, it will forecast future changes in wind speed (2014-2100) at six synoptic stations in Sistan and Baluchestan, focusing on the differences between CMIP5 and CMIP6 reports regarding wind energy. This research is the first to examine future wind characteristics in Iran using CMIP6 model outputs while comparing the performance of both CMIP5 and CMIP6 in simulating wind speed in the study area. Wind energy is highly sensitive to climate change, as future alterations in wind flow characteristics will significantly impact electricity generation potential. Therefore, understanding future climate scenarios, especially under varying global warming conditions, is vital for estimating changes in wind energy resources over the coming decades.

Method

Wind speed data were obtained from six synoptic stations operated by I.R. of Iran Meteorological Organization (IRIMO), covering the period from 1990 to 2014, with all stations maintaining continuous data records throughout this timeframe. The data from CMIP5 and CMIP6 climate models were downloaded from the respective websites. This study selected outputs from six general circulation models for the historical period (1990-2014) and the future period (2014-2100) under the emission scenarios SSP2-4.5 and SSP5-8.5 for CMIP6 and RCP4.5 and RCP8.5 for CMIP5. The ability of the CMIP6 and CMIP5 climate models to simulate historical wind speed was evaluated against observational data from Sistan and Baluchestan using statistical criteria, including bias, correlation, and standard deviation. This evaluation determined the capability and accuracy of the models and assessed the uncertainty in their wind speed simulations before applying them to future climate forecasts. A multi-model averaging approach was employed to reduce uncertainties associated with individual models, utilizing the CDFT package in RStudio for downscaling and output bias correction based on cumulative distribution function transformation.

Results and Discussion

Most CMIP models simulated wind speed effectively. The CanESM5 model in CMIP6 demonstrated improved performance compared to CMIP5, yielding results closer to observational data. In contrast, the CMCC-ESM2 and CNRM-CM6-1 models in CMIP6 were less efficient than their CMIP5 counterparts. CMIP5 indicated a decrease in wind speed, while CMIP6 suggested an increase in annual projection, although these changes were not statistically significant. The projected average wind speeds by 2100 are 3.58 m/s and 3.57 m/s for the SSP2-4.5 and SSP5-8.5 scenarios, respectively, while the RCP4.5 and RCP8.5 scenarios predict averages of 3.1 m/s and 3.2 m/s, respectively. The baseline average wind speed from observational data is 3.49 m/s. CMIP5 indicated a decrease in wind speed across all months under both scenarios, with the most significant reductions occurring in spring (-0.47 m/s) and the least in autumn (-0.29 m/s). Conversely, CMIP6 projected increases in wind speed across all seasons, with the highest increase in spring (0.49 m/s) and the lowest in summer (0.32 m/s) under the SSP2 scenario. Under the SSP5 scenario, the highest increase was observed in winter (0.43 m/s) and the lowest in summer (0.37 m/s). The annual average for CMIP5 models showed a decrease in wind speed at all stations compared to the baseline period, particularly at Khash and Zabul stations. In CMIP6, all stations except Chabahar exhibited increased wind speeds, with Chabahar recording the highest average wind speed and other stations showing minimal differences. The results indicate varying model performance in simulating climate variables, with the historical wind speed uncertainty in CMIP5 models potentially attributed to differences in grid resolution, atmospheric components, and convection scheme parameterization.

Conclusion

Given the observed biases in the models, future research should involve a comprehensive study utilizing additional models from the CMIP6 and CMIP5 families. These findings are significant for assessing the potential of the wind energy sector in Sistan and Baluchestan, a region known for its wind resources, which may inform future development strategies. It is also recommended that similar assessments be conducted in other regions of Iran to determine whether identified sites with suitable wind power may experience future resource fluctuations. Also, it is suggested that different methods of bias correction and downscaling should be investigated and the best method should be suggested. On the other hand, using satellite data instead of observational data and comparing their results can be considered as another research proposal.



Key Words:Downscaling, Wind speed, Sistan and Balochestan, CMIP models



Article Type: Research Article

Acknowledgement

We express our sincere gratitude to the University of Zanjan for their financial and logistical support throughout this research project, identified by code 1174-2-1402.

Conflicts of Interest

The authors declare no conflicts of interest regarding the authorship or publication of this article.

Data Availability Statement

Datasets are available upon reasonable request to the corresponding author.

Authors’ Contribution

Author 1: Formal analysis and investigation, Software

Author 2: Writing and manuscript editing