Solar energy, a renewable and clean source, poses challenges in predictability due to cloud interference. A recent study by Brookhaven National Laboratory, in collaboration with several institutions, has made strides in understanding how various cloud types affect solar energy forecasting. The research used data from the U.S. Department of Energy’s Atmospheric Radiation Measurement Program, spanning from 2001 to 2014.
Yangang Liu, a senior scientist at BNL, explained, “Thanks to the decade-long, high-quality data collected by the U.S. Department of Energy’s Atmospheric Radiation Measurement (ARM) Program, our study offers a detailed evaluation of solar forecasting models across varied cloud conditions.”
Researchers analyzed eight cloud types, including cumulus and stratiform, to assess their impact on solar irradiance predictions. The study leveraged physics-informed data-driven models, previously developed by the team, and tested them against real-world data from the ARM South Great Plain Central Facility site. These findings highlighted a hierarchy in model accuracy, with models performing best under weak convective clouds like cirrus, and worst with strong convective clouds such as deep convective clouds.
Shinjae Yoo, from Stony Brook University and Brookhaven Lab, commented on the categorization of cloud types: “By categorizing clouds into stratiform, weak, and strong convective types, we were able to identify where our models performed best and where they needed improvement. The trends we saw highlighted the complexity of forecasting under certain cloud conditions. For example, in the case of deep convective clouds — which have more complex spatial structures with dynamic and unpredictable nature — we noticed a significant uncertainty in the results.”
Results from this study could enhance solar forecasting by offering clearer insights into the interactions between cloud types and solar irradiance, potentially improving the accuracy and reliability of solar energy predictions.
Read the full story by Ankita Nagpal at the AI Innovation Institute website.