The China Meteorological Administration (CMA) has released Fenghe, its first AI meteorological service system. Jointly developed with Tsinghua University, it is designed to promote “the intelligent transformation” of meteorological services, the CMA said.
Fenghe comprises five core modules: the Meteorological Knowledge Center, Model Square, Meteorological AI Toolbox, Intelligent Agent Factory and Evaluation Laboratory.
According to the CMA, it can understand meteorological service requirements, generate service content, make informed inferences and decisions, and use meteorological tools.
Intelligent transformation
Users of Fenghe can enter weather or service-related queries into the app or website version to received tailored, analyzed suggestions and answers. This can be across a variety of scenarios, including public science, general weather inquires, travel and risk warning advice.
It is also applicable to fields closely linked with weather such as transportation, tourism, healthcare, logistics and energy.
1+1+N Technology framework
Fenghe has been designed with what the CMA calls a “1+1+N” technology framework – compromising a Dual-Core and Multi-Scenario.
The first core is a base model that integrates general purpose AI models with professional meteorological knowledge. Through incremental pre-training specific to the meteorological field, it has gained an accurate understanding of meteorological standards and disaster mechanisms, CMA said.
The second core is an intelligent agent development platform that integrates data interfaces, toolchains and scenario-based services. This enables the model to access real-time data from various CMA AI-enabled weather systems.
Multi-Scenario involves developing specialized application-specific intelligent agents with dedicated reasoning capabilities for various users, including the public as well as specific industries.
In related news, AI weather prediction potential was explored in a WMO-UAE workshop