A multi-model multi-method ensemble (4ME) assessment of future photovoltaic potential in northern China

To incorporate the impact of future climate change into solar resource assessment, it is necessary to consider the output of global climate models (GCMs). The GCM outputs, primarily due to the inadequate representation of physical processes within the model, display biases that require correction. However, numerous studies focusing on future changes in solar energy potential are confined to using a single bias-correction method. Furthermore, the estimation of photovoltaic (PV) power often relies on a limited set of empirical formulas, which overly simplifies the weather-to-power conversion mechanisms. To address these deficiencies, this work presents a multi-model multi-method ensemble scheme and introduces a refined approach to modeling PV systems. This research separates itself from previous works in two facets: (1) A multitude of GCMs and bias-correction methods are leveraged, and thus return a more reliable solar resource assessment than using a single model–method combination, and (2) the physical model chain is used for weather-to-power conversion, which allows refined PV modeling with better accuracy. Empirically, this work examines the spatio-temporal variations of the bias-corrected meteorological variables and PV power in northern China from 2025 to 2059 under three climate change scenarios (SSP126, SSP245, and SSP585). The results indicate that, under the SSP126 scenario, the inter-decadal change rates for annual insolation, annual average temperature, annual average wind speed, and solar capacity factor are 5.99 kWh/m2, 0.27°C, −0.011 m/s, and 0.095%, respectively. In contrast, these values are 3.79 kWh/m2, 0.64°C, −0.013 m/s, and −0.111% for the SSP585 scenario. This work delivers a refined assessment of future solar potential under climate change, which can provide valuable insights during the transition to a cleaner power grid.