AnswersPublished17 sources
What is the China Meteorological Administration’s Fengqing AI global short and medium range ensemble forecasting system (CMA AIM GFS Fengqing ENS V1.0), reported on October 17, 2026, and how does its operational review at the National Meteorological Centre and deployment for nati
CMA AIM GFS Fengqing ENS V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short and medium range probability weather forecasting, developed under the National Meteorological Centre’s leadership.[9] October 17 reports describ The 2024 CMA–Tsinghua Fengqing model incorporated atmospheric physics and ph...
Published byImages generated with GPT Image 2
Research answer
Key takeaways
- CMA AIM GFS Fengqing ENS V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short and medium range probability weather forecasting, developed under the National Meteorological Centre’s leadership.[9] October 17 reports describ
- The 2024 CMA–Tsinghua Fengqing model incorporated atmospheric physics and physical conservation properties into training to improve physical interpretability and long lead forecast behaviour.[2] However, the retrieved evidence does not subs
- Insufficient evidence in the supplied excerpts confirms the ensemble’s precise vision Transformer architecture, its particular physical constraints, or specific probability products for heavy rain, typhoons and severe convection.
- Its verified operational description is a global probability forecast system with 51 members.[9] Computing and performance October 17 reporting says training and iteration used a 100,000 card AI4S supercomputing cluster at the National Supe
AI Prompt
openai.comCreate a landscape editorial hero image for this Studio Global article: What is the China Meteorological Administration’s Fengqing AI global short and medium range ensemble forecasting system (CMA AIM GFS Fengqin. Article summary: CMA AIM GFS Fengqing ENS V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short and medium range probability weather forecasting, developed under the National Meteorological Centre’s leadership.[9] Octobe. Topic tags: general web, ai, benchmarks, anthropic, finance. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, cha
CMA-AIM-GFS-Fengqing-ENS-V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short- and medium-range probability weather forecasting, developed under the National Meteorological Centre’s leadership.9 October 17 reports described it as deployed nationally after passing operational review, but a later official CMA notice scheduled formal operational service for November 10, 2026—an important distinction between the initial announcement and the official start date.
7
9
Operational status and purpose
- The operational review took place at the National Meteorological Centre, with experts from atmospheric research, numerical forecasting, meteorological information services, regional meteorological bureaus and computing organizations; passing established that the system met the conditions for operational use.
4
- Development partners identified in the reports include the Xiong’an Meteorological AI Innovation Research Institute, National Meteorological Information Centre and National Supercomputing Internet, alongside the National Meteorological Centre.
9
- CMA’s later specification describes twice-daily global forecasts covering 0–360 hours, at six-hour intervals and 0.25° horizontal resolution, with 51 ensemble members and upper-air and surface weather variables.
How it differs from the earlier Fengqing model
- Conceptually, a deterministic forecast gives one predicted atmospheric evolution; an ensemble supplies multiple plausible evolutions, allowing their differences to represent uncertainty and support probability estimates rather than just one forecast.
- The 2024 CMA–Tsinghua Fengqing model incorporated atmospheric physics and physical conservation properties into training to improve physical interpretability and long-lead forecast behaviour.
2 However, the retrieved evidence does not substantiate the specific approximately 3-billion-parameter, 25-kilometre, 15-day-in-three-minutes specifications, so those should not be transferred to the ensemble version as verified specifications.
- Insufficient evidence in the supplied excerpts confirms the ensemble’s precise vision Transformer architecture, its particular physical constraints, or specific probability products for heavy rain, typhoons and severe convection. Its verified operational description is a global probability-forecast system with 51 members.
Computing and performance
- October 17 reporting says training and iteration used a 100,000-card AI4S supercomputing cluster at the National Supercomputing Centre in Tianjin, and describes the system’s data, model and computing as domestically autonomous and controllable.
7 That establishes a reported domestic-control claim, not a documented audit of every data source or component.
- Testing on independent samples from January 2025 through August 2026 reportedly yielded an 11.0-day effective forecast range for ensemble-mean global 500 hPa geopotential height, with favourable continuous ranked probability score (CRPS) and spread–error ratio results.
8 This is a skill result for a specified atmospheric variable, not a guarantee that all weather hazards are accurately predictable eleven days ahead.
- “Independent samples” should not be equated with independently replicated verification: the supplied excerpts do not establish an external, reproducible evaluation.
- Reports claim overall superiority to ECMWF’s AI ensemble, but the available evidence does not provide a detailed, matched comparison against both ECMWF’s numerical ensemble and AIFS-CRPS.
16 Insufficient evidence supports treating that claim as established across variables, regions and lead times.
- The supplied evidence does not verify an exact minutes-per-global-run inference benchmark for the ensemble. The reported large training cluster should therefore not be interpreted as the computing requirement for each subsequent forecast.
Intended use and remaining uncertainty
- The reporting presents extreme-weather warning as a future role, while CMA describes the system as supporting global operational probability forecasts.
- The clearest supported conclusion is that Fengqing Ensemble adds operational uncertainty forecasting to CMA’s AI capabilities; its formal deployment timetable is clarified by the later official CMA notice, while detailed architecture, runtime and comparative performance claims remain insufficiently documented in the evidence available here.
Studio Global AI
Continue your research
This page includes a source-backed answer you can continue inside Studio Global.
People also ask
What is the short answer to "What is the China Meteorological Administration’s Fengqing AI global short and medium range ensemble forecasting system (CMA AIM GFS Fengqing ENS V1.0), reported on October 17, 2026, and how does its operational review at the National Meteorological Centre and deployment for nati"?
CMA AIM GFS Fengqing ENS V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short and medium range probability weather forecasting, developed under the National Meteorological Centre’s leadership.[9] October 17 reports describ
What are the key points to validate first?
CMA AIM GFS Fengqing ENS V1.0, or “Fengqing Ensemble V1.0,” is CMA’s AI system for global short and medium range probability weather forecasting, developed under the National Meteorological Centre’s leadership.[9] October 17 reports describ The 2024 CMA–Tsinghua Fengqing model incorporated atmospheric physics and physical conservation properties into training to improve physical interpretability and long lead forecast behaviour.[2] However, the retrieved evidence does not subs
What should I do next in practice?
Insufficient evidence in the supplied excerpts confirms the ensemble’s precise vision Transformer architecture, its particular physical constraints, or specific probability products for heavy rain, typhoons and severe convection.