A high-resolution downscaling approach for solar irradiance using statistical parameter matching

The limited intra-hour variability of globally available hourly renewable energy system data leads to inaccuracies in the modeling of renewable energy systems. While sub-hourly data can improve model accuracy, such data are not globally available. The existing approaches to increase the temporal resolution of solar irradiance often rely on site specific measurements or complex models, limiting global scalability. This work, therefore, presents a methodology to increase the temporal resolution of the global horizontal irradiance from 1 h to 1 min using non-dimensional irradiance and parameters matching based on daily irradiance characteristics for arbitrary locations. The methodology is validated using statistical methods and energy system optimization. The hourly annual normalized root mean square error and Kolmogorov-Smirnov Integral range from 5 to 7 % and 0.1–0.7, respectively, for different locations consisting of varying weather conditions. The energy system optimization results of the synthetic data demonstrate superiority in terms of cost and feasibility relative to the average hourly resolution data. The use of synthetic minute resolution data significantly improves the design accuracy of dynamic components such as inverters and storage systems. The globally applicable method, based on Köppen-Geiger classification coverage, will enable more reliable energy systems modeling in the future.