#China #Shanghai #Envision #Tianji #AI Weather Quest

Envision's Meteorological Triumph at AI Weather Quest

In a landmark achievement for AI-driven meteorological forecasting, Envision, a global leader in green technology, recently announced that its independently developed weather foundation model, Tianji, clinched the top position at the AI Weather Quest. Organized by the European Centre for Medium-Range Weather Forecasts (ECMWF), this prestigious competition pits the world's leading forecasting teams against each other, testing their capabilities in predicting weather patterns using cutting-edge artificial intelligence.

Tianji emerged victorious by outperforming nearly 20 elite teams across various critical indicators. The metrics evaluated included average variability, near-surface temperature, mean sea-level pressure, and weekly precipitation forecasts for a prediction window stretching from 19 to 25 days ahead. The AI Weather Quest continuously evaluates models through real-time forecasting and verification, assessing their performance in diverse weather conditions.

This competition is distinct as it focuses on probabilistic forecasts, emphasizing more than traditional parameters. It challenges participants to forecast average weekly temperatures and mean sea-level pressure while also predicting accumulative weekly precipitation. The evaluation process scrutinizes grid cells predominantly over land for temperature and precipitation, while the mean sea-level pressure is assessed globally.

The success of Tianji comes on the heels of a series of impressive accolades. It ranked at the highest level in the global WeatherBench 2 evaluation alongside Google's WeatherNext2 model and received the first place in the overall rankings within the deterministic forecasting trajectory of the meteorological AI model demonstration program from China's National Meteorological Center. Tianji has also proven its effectiveness in energy sectors, winning first place in AI load forecasting competitions by State Grid and the China Southern Power Grid. This recent triumph reinforces Envision's assertion of Tianji's robust capabilities across forecasting ranges, from short-term to subseasonal periods.

Robert Zheng, head of Envision's Tianji weather foundation model, remarked: "Subseasonal forecasts, ranging from two weeks to two months, represent one of the most formidable challenges in meteorology. The longer the prediction horizon, the greater the uncertainty regarding atmospheric changes, making it crucial for the model to detect evolving weather trends effectively."

Drawing on years of practical experience in new energy, Envision has pioneered the first weather model worldwide that seamlessly integrates physics with AI. Utilizing cutting-edge spatio-temporal high-definition data, at a petabyte scale, along with a three-dimensional space-air-ground data network, Tianji provides reliable forecasts across all time scales. Enhanced by high-performance computing, Tianji facilitates rapid, global short-term forecasts with significantly improved accuracy concerning wind speed and solar radiation in strategic areas, making it an invaluable tool for real-time monitoring and long-term planning.

With its extensive forecasting capabilities, Envision’s Tianji stands at the forefront of meteorological innovation, setting new standards that could transform not just how weather is predicted but also its implications for related sectors including energy and environmental management. As the world grapples with the challenges posed by climate change, such advancements in weather forecasting technology have never been more vital. Envision's achievement in AI Weather Quest is not only a testament to their technological prowess but also a promising step towards more accurate and reliable weather predictions for the future.