Utilizing PV output for retrieving cloud information

Clouds are a key component in weather and climate, influencing both incoming solar shortwave radiation and outgoing thermal radiation. During recent years, electricity production using solar photovoltaic (PV) panels has grown rapidly worldwide. As the number of PV installations continues to grow, it is apparent that the network of PV installations constitutes a highly interesting, potential new source of cloud information. From a meteorological perspective, there is a connection between solar electricity production (PV output), solar radiation and prevailing cloud conditions. When the meteorological conditions are known, the electricity production of a known PV system can be accurately modeled. Here, the cloud optical depth (COD), a parameter of the cloud optical properties, is of central importance, as it governs how incoming solar radiation attenuates due to clouds.

In this study, we have developed a new, fast, accurate and universally applicable physically-based approach for deriving COD directly from PV output measurements. In addition to these latter, the approach uses atmospheric variables such as wind speed, air temperature, and cloud-free solar radiation components altogether produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) as well as an overcast sky selection algorithm. The study is carried out at two stations of the Finnish Meteorological Institute providing relevant ground-based measurements. The approach exhibits a similar or better performance than an earlier developed and published state-of-the-art method when compared to ground-based and satellite-based COD retrievals serving as reference. It is intended to apply this approach over other locations in the world where PV output measurements are available.