Development of an advection model for solar forecasting based on ground data. Part II: Verification of the forecasting model over a wide geographical area

•A new solar forecasting model based on ground data is proposed.•An advection model of cloud dynamics is developed.•The advection model is verified for broad geographical areas.•The model is simple, physically meaningful, and in good agreement with.•measurements.•The model clearly relates to satellite forecasting and machine learning forecasting. A new spatiotemporal forecasting model for photovoltaic output power, outlined in the first paper, is validated herein against irradiance measurements. Irradiance measurements were first interpolated onto a grid. By repeating this process over several time steps, cloud motion was visualized, and the velocity of the clouds was estimated. Using these estimations of cloud velocity, the cloud motion for a future time interval was simulated using advection-diffusion equations and the average irradiation over a target region was estimated. The model was applied to a 170 km × 60 km region in northern Kyushu, Japan. Predictions of the transients for average irradiation were in good agreement with ground measurements. Simulation of the distribution of irradiation within the photovoltaic system also agreed well with ground measurements. This approach suggests an additional function for photovoltaic power systems, not only as a power source but also as a sensor for weather forecasting. This application provides a clear incentive to install more photovoltaic power systems.