Temporal upscaling of solar radiation components using an analytical model for variability modeling

To properly design systems that harness the solar resource, reliable measurements or estimates of its availability are needed. Often, this data is only available at temporal resolutions that are not sufficiently fine to capture the short-term fluctuations of the solar resource, creating uncertainty in the design and performance assessment of solar-powered systems. Synthetic irradiance generation techniques aim to create high temporal resolution estimates starting from lower resolution data, however, a literature review on the topic shows opportunities for improvement regarding the generality of the generation and evaluation approaches. This work proposes novel methods to estimate a quantifiable indicator of short-term variability from hourly data, a new temporal upscaling methodology for time series of arbitrary origin and destination temporal resolution, based on explicit mathematical functions and sufficient generality for application in different contexts, and an improved assessment methodology. The method was applied tor ground and satellite upscaling up to 1 min, resulting in time series with characteristics consistent with those of a higher temporal resolution and exhibit minimum deviations against the original time series characteristics. This method has applications not only for upscaling of low temporal resolution data, but also for gap-filling techniques and the upscaling of hourly solar radiation forecasts.