Accurate intraday solar forecasts are crucial for electricity trading and microgrid management. While satellite-based methods outperform numerical weather prediction (NWP) models for short-term horizons, their accuracy depends heavily on stable weather conditions, performing poorly during convection, fog, or large depressions. This study investigates the impact of various North Atlantic weather regimes—Atlantic Ridge, Scandinavian Blocking, NAO+, and NAO-—on the reliability of satellite-based forecasts.

We conducted an 8-year backtest using forecasts generated four hours ahead with a 15-minute time step and validated them against pyranometer data. Our analysis shows that forecast errors vary significantly with the prevailing weather regime. The difference in relative RMSE between the Scandinavian Blocking and Atlantic Ridge regimes was 10-12% in summer (2016-2020) and approximately 10% after 2020. In winter, this difference was around 20% before 2020 and 15% after 2020. These findings demonstrate that large-scale atmospheric patterns significantly influence forecast reliability. Given that weather regimes can be predicted in advance, this analysis provides valuable insights for anticipating forecast error, which can help optimize PV integration and serve as a useful input for machine-learning-based forecast algorithms.

These variations in weather regime frequencies directly impact forecast errors, emphasizing the importance of large-scale atmospheric patterns in solar energy forecast reliability. As weather regimes can be predicted several days in advance, this analysis provides useful information to anticipate the magnitude of forecast error and therefore adapt suitable decisions for optimizing PV integration management and electricity trading and provide important insights to develop deep learning forecast algorithms