/PRNewswire/ –Envision, a global leader in green technologies, today announced that its independently developed Weather Foundation Model Envision Tianji took first place overall in the latest edition of the AI Weather Quest, an AI weather forecasting competition organized by the European Center for Medium-Range Weather Forecasts (ECMWF). Tianji outperformed nearly 20 top teams worldwide, ranking first in all categories assessed, including variable average ground temperature, average sea level pressure and precipitation for the 19-25 day forecast period.
The AIWeather Quest evaluates models through continuous real-time forecasting and verification and assesses their sustained performance under various weather conditions. Beyond the 15-day medium-range forecast horizon, the competition focuses on probabilistic forecasts of weekly average temperature and mean sea level pressure and weekly precipitation for days 19-25. Temperature and precipitation are evaluated across predominantly land-covered grid cells, while mean sea level pressure is assessed globally, emphasizing the ability to predict weekly conditions and their uncertainty several weeks in advance. The teams submit weekly weather forecasts, which are compared by the organizers with the actual weather data and then publicly updated in the rankings.
This win follows a series of top achievements for Tianji, including a top ranking in the global WeatherBench 2 assessment, on par with the Google WeatherNext2 model, and first place overall in deterministic forecasting from the AI Weather Forecast Model Demonstration Program organized by China’s National Meteorological Center. Tianji has also demonstrated its value in the field of energy applications, winning first place in State Grid’s AI load forecasting competition and China Southern Power Grid’s business competition. With this latest result, Envision has once again demonstrated that Tianji covers the entire spectrum from short-term forecasts to forecasts for periods of several weeks (so-called subseasonal forecasts).
“Subseasonal forecasting – two weeks to two months in advance – is one of the biggest challenges in meteorology,” said Robert Zheng, head of the Tianji Weather Foundation Model at Envision. “The longer the time horizon, the greater the uncertainty in atmospheric changes and the greater the demands on a model’s ability to detect evolving weather trends. Building on Envision’s years of practical experience in new energy, we have developed the world’s first physics-based AI weather model that leverages petabyte-scale high-resolution spatial and temporal data and a three-dimensional data network from space, air and ground to provide reliable forecasts over the entire period deliver.”
Thanks to powerful computer technology, Tianji delivers fast short-term and global forecasts with significantly improved accuracy for wind speed and solar radiation in key regions. It covers the entire period from minute-by-minute forecasts to subseasonal trend analysis, providing comprehensive meteorological information for real-time monitoring and long-term planning.
