Statistical downscaling of climate models projection of minimum temperature under RCP scenarios in Western of Iran

Considering the significant effect of chilling and frost phenomenon on agricultural production in Iran, the aim of this study is projection of minimum temperature in three provinces of Kurdistan, Kermanshah and Ilam, Western of Iran. For this purpose, the data of 17 meteorological stations during the baseline period of (1989-2018) were collected. Then, the HadGEM2 and CanESM2 climate models outputs were statistically downscaled using LARS-WG and SDSM weather generators under three climate change scenarios of RCP2.6, RCP4.5 and RCP8.5 during future period (2021–2050) and compared with observed data. The performance of the two weather generators, were compared using MSE, RMSE, MAE and R2 indices. The results indicated the good accuracy of both statistical models in simulating the minimum temperature in the study area; however the SDSM model performed better than the LARS-WG. The projected changes of minimum temperature compared to the baseline period revealed a significant increase varying between 0.6 – 1.5 oC in study stations. The most significant change was observed in northern parts of study region especially in the Saqhez and Zarineh stations. The findings of this study can be used in frost risk management and agroclimatic planning in the region.