Assessment of RegCM4 model for estimation of total solar radiation (Case study: Chaharmahal and Bakhtiari province)

The amount of energy received from the sun on the ground needed to estimate crop water use, use of clean energy and issues related to climatic factors. In this study, the ability of two estimation methods based on statistical and dynamic methods were evaluated. Angstrom, as a widely accepted statistical model, and Regional Climate Model (RegCM version 4.0) were chosen to obtain radiation estimations. To perform this study, meteorological data of 6 IRIMO’s automatic weather stations located in Chaharmahal and Bakhtiari province, west of Iran, were used. RegCM4 model was run considering initial values and boundary conditions of NNRP1 ,at 18 sigmoid levels, with spatial resolution of 30 km and temporal resolution of 150 minutes, in the period of 2010 to 2014. Results showed the RegCM4 model output has the minimum Root mean square error (RMSE) of 10 W m-2 in warm months (June to September) and highest RMSE of 408 W m-2 in February. The lowest RMSE was obtained in September in Borujen station and the highest value was recorded in month of February in Ardal. The average value of overall RMSE in the region was 175 W m-2. Monthly values of Angstrom model was calibrated for study stations. The highest correlation coefficient and lowest RMSE and MBE were obtained for September. The maximum value of RMS was 321 W m-2 in April. Angstrom model showed an underestimation in 73% of the cases while RegCM showed an overestimation 93 % of the cases. These results tend to confirm the need for post-processing of the climate model outputs.