A new climatology reference model to benchmark probabilistic solar forecasts. (15th July 2021)

• Record Type: Journal Article
• Authors: Le Gal La Salle, Josselin; David, Mathieu; Lauret, Philippe
• Is Part Of: Solar energy. Volume 223(2021)
• Page Start: 398
• Page End: 414
• Publication Date: 2021-07-15
• Subjects: Benchmarking -- Solar irradiance -- Probabilistic forecasting -- Climatology -- Reference model; Solar energy -- Periodicals; Solar engines -- Periodicals; 621.47
• DOI: 10.1016/j.solener.2021.05.037
• Languages: English
• ISSNs: 0038-092X

Abstract

Highlights: A new reference model for probabilistic solar forecasts called "CSD-CLIM" is introduced. There is no need to form probability distributions to compute its CRPS. A baseline for climatology benchmark models is proposed and used to compare models. CSD-CLIM is easy-to-implement and meets all prerequisite properties of a good benchmark model. CSD-CLIM achieves the best trade-off between reliability and resolution. It can be a viable alternative to existing reference models. Abstract: Probabilistic solar forecasting is becoming a major topic in the solar research community as it provides more information about the uncertainty of the forecast compared to deterministic forecasting. However, to facilitate the adoption of probabilistic forecasts within solar forecasting communities (industry and academic), the definition and the use of standardized best practices are a prerequisite. Among others, there is a need for benchmark models that are able to properly assess the performance of new probabilistic forecasting methods. In this work, we propose a new climatology benchmark model called "CSD-CLIM" (for Clear-Sky Dependent Climatology). This new reference model is evaluated against two other climatology benchmark models namely the naive climatology and a well-referenced model in the literature, the CH-PeEn (for Complete History Persistence Ensemble). The verification of compliance with a set of properties that a climatology benchmark model must follow demonstrates that the new CSD-CLIM model outperforms the naive climatology and that it can be a viable alternative to the CH-PeEn model. It is shown that the better performance of CSD-CLIM is due to a specific binning of the historical irradiance data based on the clear-sky irradiance values.