Methodology to Stochastically Generate Synthetic 1-Minute Irradiance Time-Series Derived from Mean Hourly Weather Observational Data
Abstract
Well geographically distributed high temporal resolution solar irradiance data is scarce, resulting in many studies using mean hourly irradiance time-series as an input. This research demonstrates that by taking readily available mean hourly meteorological observations of okta, wind speed, cloud height and atmospheric pressure; 1-minute resolution irradiance time-series that vary on a spatial dimension can be produced. The synthetic time-series temporally validates against observed 1-minute UK irradiance data with 99% K-S test confidence levels across 3 metrics of variability indices, ramp-rate occurrences and irradiance frequency. A new methodology is applied to existing research that produces two-dimensional cloud cover using a vector approach to add spatial correlation to irradiance time-series, as well as improvements to the clear-sky index calculations. The methodology is applied to a hypothetical configuration to demonstrate its capabilities.
Keywords
Irradiance generation, Minute resolution, Resource modelling, Cloud edge enhancement.