Spatiotemporal differences and uncertainties in projections of precipitation and temperature in South Korea from CMIP6 and CMIP5 general circulation model s

Abstract This study compared the historical simulations and future projections of precipitation and temperature of Coupled Model Intercomparison Project (CMIP)5 and CMIP6 general circulation models (GCMs) to quantify the differences in the projections due to differences in scenarios. Five performance indicators were used to quantify the model reproducibility of the observed precipitation levels at 22 stations for the historical period of 1970–2005. The percentages of change in precipitation and temperature were estimated for the near (2025–2060) and far future (2065–2100) for two Representative Concentration Pathway (RCP)4.5 and RCP8.5 scenarios of CMIP5 and two Shared Socioeconomic Pathway (SSP)2–4.5 and SSP5‐8.5 scenarios of CMIP6. The uncertainty in the projection in each case was calculated using the reliability ensemble average (REA) method. As a result, the CMIP6 GCMs showed an improvement compared with the CMIP5 GCMs with regard to the ability to simulate the historical climate. The uncertainty in the precipitation projections was higher for SSPs than that in RCPs. With regard to the temperature, the uncertainty was higher for RCPs than for SSPs. The ensemble means of the precipitation and temperature showed higher changes in the far future compared with the near future for both RCPs and SSPs. This study contributes to improvement in the confidence of future projections using CMIP6 GCMs and bolsters our understanding of the relative uncertainty in SSPs and RCPs.