Characterizing solar radiation based on frequency amplitude and variability

Understanding atmospheric variability, the physical properties of transitional cloud cover and interdependence of solar radiation is critical in solar radiation modelling and forecasting. Lewis et al. propose a novel 15-minute period solar resource classification algorithm based on a time to frequency domain transform, which requires minimal inputs.

This method allows the researchers to explore solar radiation in the frequency space, which provides the necessary dissociation from the nature of the atmosphere and solar radiation while providing access to the underlying process of the solar radiation time series.

“Our resulting theory is that the characterization of solar radiation depends on two measures, namely the measure of amplitude and variability,” said author Carmen Lewis.

In other words, the researchers could determine the conditions under which solar radiation is measured by considering the amplitude and variability of the converted solar radiation frequency series.

Solar resource classification typically considers the measured amount of solar radiation on the Earth’s surface to determine the representative sky conditions, such as whether there are clouds in the sky dome.

Many solar resource classification models are dependent on geographical location either through a threshold or clear-sky model that requires previous establishment and consolidation. This means, however, that these models have limited application in data-scarce locations, such as southern Africa.

“I think this project has shown that there are simple solutions to model relatively complex processes,” said Lewis. “It is critical that solutions emerge for these conditions to enable wider adoption of renewable energies as competitor energy generation technologies.”

Source: “A solar resource classification algorithm for global horizontal irradiance time series based on frequency domain analysis,” by C. Lewis, J. M. Strauss, and A. J. Rix, Journal of Renewable and Sustainable Energy, (2021). The article can be accessed at https://doi.org/10.1063/5.0045032 .