Assessment of new solar radiation nowcasting methods based on sky-camera and satellite imagery

This work proposes and evaluates methods for extending the forecasting horizon of all-sky imager (ASI)-based solar radiation nowcasts and estimating the uncertainty of these predictions. In addition, we evaluated procedures for improving the temporal resolution and latency of satellite-imagery-derived solar nowcasts. Based on these contributions, we assessed the reliability of ASIs and satellite-derived solar radiation nowcasts, with 1-min time-resolution and up-to-90-min ahead. The study was conducted in a location in Southern Spain using a set of cloudy days, specifically selected as representative of the most challenging conditions regarding solar radiation nowcasting. The results reveal that the use of ASI-based models provide low benefits compared to the use of satellite-based models for point solar radiation nowcasting. Given the frequency of occurrence of the different sky types in the study area, the results suggest that the use of a simple smart persistence algorithm, in combination with a low-resolution satellite nowcasting model could be an adequate choice, avoiding the challenges associated with the use of ASIs.

Automated summary

This study proposes and evaluates methods for extending the forecasting horizon of all-sky imager (ASI)-based solar radiation nowcasts and improving the temporal resolution and latency of satellite-imagery-derived solar nowcasts. The results suggest that the use of ASI-based models provide low benefits compared to satellite-based models for point solar radiation nowcasting, and recommend the use of a simple smart persistence algorithm in combination with a low-resolution satellite nowcasting model based on the frequency of occurrence of different sky types in the study area.