AI bringing innovation to Earth system prediction
The Early Warnings for All initiative is dedicated to ensuring protection against hazardous weather, water or climate events for everyone on the planet. WMO is committed to leveraging new technologies, including Artificial Intelligence (AI), to enhance Earth system observations and predictions.
Given the significant and rapid advancements in AI research and application within the meteorological community, even in the operational context, AI technologies have great potential to enhance the capability of WMO Members in Earth system observation and prediction in the near future.
In this INFCOM-3 side event, the latest progress on the use of AI technologies was presented, with a focus on data-driven forecasts, satellite nowcasting and post-processing of products.
- Mr David Richardson (ECMWF) presented an overview of recent progress in data-driven weather forecasting, illustrating how rapidly this field is advancing. He also noted areas where further development is still needed such as understanding extreme events and representation of forecast uncertainty.
- Dr David John Gagne (NCAR, USA) introduced a new pilot project on AI-based satellite nowcasting to show the benefit of using AI to derive new nowcasting fields from the vast amount of information in the satellite data. AI-based nowcasting products will serve participants from RAI, RAII, and RAIII to enhance their early warning service in this pilot project.
Dr Kan Dai and Dr Bo Lu (both from CMA, China) presented the latest development of AI-based post processing applications in their country. Typical NWP post-processing scenarios, including model bias correction, forecasts downscaling, multi-data/model fusion, hazardous weather warning, and data-driven AI models, were presented.
For daily highlights, visit the INFCOM-3 dedicated webpage.