"风云"卫星俯瞰全球，筑就陆海空天一体化监测体系；人工智能预报精准至公里级，暴雨预警准确率再创新高；生活气象服务指数增至70余 种，精细化气象服务涵盖衣食住行游购娱……"十四五"时期，从守护百姓衣食住行到助力国家发展大局，从筑牢国内生产生活安全防线到赋能 产业升级，气象服务已成为满足民生期待、推动经济社会发展的重要支撑。 跟随这组海报，共同感知中国气象以硬核实力护航国计民生、守护万家平安、赋能千行百业的蓬勃力量与责任担当。 · 全球首次实现气象卫星区域250米分辨率每分钟连续成像 观测 · 风云三号G星是我国首颗、全球第三颗低倾角轨道降水测量 上 • 风云三号E星是全球唯一在晨昏时段开展观测的民用极轨 气象卫星 新华社 司刊 nlist 新华社 3 27 建成临海空 本省合量量表机机系统 · 风云卫星、北斗探空、天气雷达等探测装备、技术实现 高水平自立自强 · "风雷""风清""风顺""风宇"气象预报模型 达到国际先进水平 · 远洋气象导航、全球碳源汇监测核枝等一批创新技术成果涌现 部信息 精准频报 年47周三日可控的全球、区域数国预报技术的 · 灾害性天气监测率提升到83% 强对流天气预警时间提前13% · ...

Core Viewpoint - The rapid development of artificial intelligence (AI) technology is enhancing the accuracy and efficiency of weather and climate predictions, with AI models expected to match or exceed traditional physical models in both temporal and spatial resolution [1]. Group 1: AI Weather Models - AI weather models, such as Pangu, Fengwu, and Fuxi, are increasingly being utilized in typhoon prediction and short-term climate forecasting, showcasing their potential to improve prediction accuracy and efficiency [2]. - The advantages of AI weather models include faster computation times, higher accuracy in forecasts (e.g., typhoon path predictions), and built-in optimization modules for various predictive studies [2][3]. - AI models have successfully predicted the El Niño-Southern Oscillation (ENSO) phenomenon for over 18 months, significantly surpassing traditional models [2]. Group 2: Integration of AI and Traditional Methods - Current AI weather models face challenges, such as unclear physical mechanisms behind predictions and insufficient temporal resolution, with some models refreshing data only every six hours compared to traditional models that do so in minutes [4]. - The integration of high-performance computing (HPC) with AI is proposed as a solution to enhance prediction accuracy and overcome computational limitations [4][5]. - Future developments in weather forecasting are expected to rely on a combination of HPC and AI, requiring advanced supercomputing systems capable of handling massive computational demands [5].