Researchers from the American Institute of Physics created a method that estimates the amount of sunlight reaching solar panels and predicts cloud cover. The method is called Spectral Cloud Optical Property Estimation (SCOPE). A solar energy plant's output is highly dependent on cloud cover.

Light coming from the Earth's surface and detected by the Advanced Baseline Imager (ABI) aboard the GOES-R satellite, shown as a function of wavelength. Source: Carlos F.M. Coimbra

Weather forecasting can predict the amount of sunlight reaching the panels, but predicting cloud cover is harder because it is characterized simply by cloudy, partly cloudy or clear. Weather forecasting systems don’t provide accurate information for estimating the amount of sunlight that is available.

In 2016, NASA launched a new generation of Geostationary Operational Environmental Satellites (GOES-R) series. These satellites occupy fixed positions above the Earth’s surface and are fixed with sophisticated instruments including the Advanced Baseline Imager (ARI). ARI detects radiation and upwelling from the Earth at specific wavelengths. SCOPE uses data from these satellites.

The SCOPE method estimates three properties of clouds that determine the amount of sunlight reaching the Earth’s surface. The first property is cloud top height, the altitude corresponding to the top of the cloud. Second is cloud thickness, the difference in altitude from a cloud’s top to bottom. The third and final property is cloud optical depth, the measure of how a cloud modifieds light passing through it. Clouds hold a lot of water in multiple forms that absorb light in different amounts, which affects optical depth.

The amount of light a cloud absorbs is dependent on the light’s wavelength. Absorption is variable for light in a wider infrared range but not variable in the narrow visible light range. The SCOPE method estimates cloud thickness, top height and optical depth with the ABI sensor data from GOES-R satellites and atmospheric models. It also uses ambient temperature and relative humidity data on the ground from ground-based weather stations. These inputs are used to adjust the temperature and gas concentration vertical profiles in the model.

The accuracy of the estimated cloud optical properties was evaluated using one year of data from 2018. Measurements were taken at seven ground-based locations in the U.S. during day and night in varying weather conditions. SCOPE can be used during the day or night with reliable accuracy.

A paper on this new method was published in the Journal of Renewable and Sustainable Energy.