Since 2013 work on building an improved solar forecasting system has been underway by researchers from the U.S. National Centre for Atmospheric Research (NCAR), located in our own Boulder Colorado.
Now, the team have unveiled their cutting-edge Sun4Cast technology; which has the potential to save the US solar industry hundreds of millions of dollars by improving predictions of clouds and other atmospheric conditions that adversely affect the output of solar panels.
Research behind the technology has been highlighted in more 20 peer-reviewed papers and the system put through comprehensive testing across states with widely varying weather conditions; including Long Island, New York; the Colorado mountains; and coastal California.
Sun4Cast has proved to be 50 percent more accurate than the solar forecasting options currently available for utilities deploying solar power. Analysis by NCAR economist Jeffrey Lazo shows utilities across the United States could save roughly $455 million as the use of solar grows.
In order to accurately predict solar irradiation levels, the Sun4Cast system relies on a vast array of observational tools, including satellite-and-sky imaging, weather radars, tailored computer software and artificial intelligence-driven mathematics. A key driver of the platform is a new atmospheric model that simulates solar irradiance based on meteorological conditions.