Intra-hour DNI forecasting based on cloud tracking image analysis

We describe an image processing methodology using Total Sky Imagers (TSIs) to generate short-term forecasts of Direct Normal Irradiance (DNI) at the ground level. Specifically, we are interested in forecasting 1-min averaged DNI values for time horizons varying from 3 to 15 min. This work describes several sky image processing techniques relevant to solar forecasting, including velocity field calculations, spatial transformation of the images, and cloud classification. The solar forecasts derived here are analyzed and quantified in terms of root mean square error (RMSE) deviations in relation to actual values and compared to the performance of persistence models.

► Novel image processing techniques for DNI forecasting are introduced.
► Comparison is done with persistence models.
► Use of sector technique for cloud decision yields short-term forecasts 20–40% better than persistence.
► Methodology is useful for time horizons between 3 and 15 min.