Researchers Min Chen and Liangchen Guo of the Institute of Urban Meteorology at the China Meteorological Administration have built a 0-6 hour ultra-short-term global horizontal irradiance forecasting system for China by combining data from the Fengyun-4B satellite with the regional numerical forecast model RMAPS-ST. Their work appears in Atmospheric and Oceanic Science Letters in a paper titled "Ultra-short-term global horizontal irradiance forecasting in China based on satellite and numerical weather prediction models."
The group tested two methods to initialize clouds in the model. In a passive mode, the system considers only advection and diffusion of hydrometeors, while in an active mode it nudges hydrometeors during the first forecast hour.
Both strategies cut forecast errors by more than 7 percent in the first 15 minutes of prediction, with the passive mode giving the best performance within the first 4 hours. Error reductions reach about 9.41 percent in summer on average, and some regions such as Sichuan see improvements of more than 20 percent.
Corresponding author Min Chen noted that, "By correcting initial cloud fields with real-time satellite data, we essentially gave the forecast system a 'live cloud map'. This is not only a technical breakthrough but also provides a practical tool for grid scheduling and China's dual-carbon strategy."
The current system performs best in warm seasons and for short lead times, while forecast errors grow as the range increases. Future development will add time-varying aerosol data and artificial intelligence methods to improve representation of cloud-radiation interactions, with the aim of supporting independent and high-precision solar power forecasting across China.
Research Report:Ultra-short-term global horizontal irradiance forecasting in China based on satellite and numerical weather prediction models
Related Links
Institute of Atmospheric Physics Chinese Academy of Sciences
Solar Science News at SpaceDaily
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |