Projecting the evolutionary path of China’s photovoltaic potential using CMIP climate models

This study provides a national-scale projection of China’s photovoltaic (PV) potential by combining CMIP6 model accuracy assessment with long-term turning-point detection. Based on rate-of-change evaluation against historical observations, four models (ACCESS-CM2, ACCESS-ESM1.5, IPSL-CM6A-LR, and KIOST-ESM) are identified as most reliable, and their ensemble mean (MM4) is used to examine PV potential from 1984 to 2100. Under low- and medium-emission scenarios, PV potential shows sustained growth with turning points around 2034 and 2028. In contrast, under the high-emission SSP585 scenario, MM4 reveals a critical turning point in 2035 followed by a multi-decadal decline until 2094, driven by a dual temperature-induced constraint: direct reductions in PV module efficiency and indirect suppression of surface irradiance through enhanced water vapor and cloud optical effects. These radiative feedbacks ultimately dominate the long-term trajectory of PV potential under SSP585. Compared with MM4, the remaining models (MM11) yield more optimistic projections but likely overestimate future PV potential. Overall, the identified turning points reflect a shift from aerosol-driven brightening to warming-driven declines, underscoring the need for high-accuracy models and adaptive strategies under high-emission pathways.