The rapid growth in the share of photovoltaic (PV) power in the electricity grid means that energy generation is becoming increasingly weather-dependent. Accurate forecasts of renewable power generation under rapidly varying conditions will therefore be indispensable for the energy grid of the future. On the other hand, weather models have limited accuracy when it comes to forecasting the optical properties of the atmosphere, and their spatio-temporal resolution is often too coarse to capture the rapid fluctuations in solar irradiance.

This seminar reports first results from an algorithm developed to infer global horizontal irradiance as well as atmospheric optical properties such as aerosol and cloud optical depth from PV power measurements. In this way the solar panels themselves are used as sensors to characterise global irradiance, and once these data are assimilated back into weather models this should lead to improvements in irradiance and PV power forecasts.

As part of the MetPVNet project, an improved forward model of PV power as a function of atmospheric conditions was developed and calibrated using PV power data from twenty systems in the Allgäu region, during two measurement campaigns in autumn 2018 and summer 2019. Irradiance, aerosol and cloud optical depth was then inferred using the radiative transfer code libRadtran, and the results compared to data from weather models and satellites. The potential of the method to extract irradiance data over a larger area as well as the improvements to irradiance forecasts from assimilating these data into the ICON weather model will be explored in future work.