浅谈技术革新对大气科学发展的推动——以数值天气预报为例  

An Insight on the Promotion of Technical Innovation to the Progress of Atmospheric Sciences:Taking Numerical Weather Prediction as an Example

作　　者：薛建军[1] 吴灿[1] 马杰 魏晓敏 Xue Jianjun;Wu Can;Ma Jie;Wei Xiaomin(China Meteorological Administration Training Centre,Beijing 100081)

机构地区：[1]中国气象局气象干部培训学院,北京100081

出　　处：《气象科技进展》2024年第5期15-21,共7页Advances in Meteorological Science and Technology

基　　金：中国气象事业发展咨询委员会研究项目“提升我国气象科技核心能力的国际比较研究”(2023--2024年);国家自然科学基金项目(42342025);中国气象局青年创新团队(CMA2023QN10)。

摘　　要：随着电子计算机、卫星遥感观测、通信网络等技术的持续进步与革新,数值天气预报从百年前的理论构想得以照进现实并不断发展。作为数值天气预报两大革命性进展,资料同化和集合预报的发展既得益于数值天气预报相关交叉领域的技术进步,同时又在技术和理念等方面推动了数值天气预报的百年发展。回望过去,随着技术革新,以统计和经验为主的天气图方法无法满足现代天气预报发展时,数值天气预报应运而生;着眼当下,人工智能助力科学探索的新范式展现出巨大的技术优势,在短时临近预报、气候预测和气象大模型领域迅速涌现出积极进展;放眼未来,我们应积极拥抱技术变革,科学看待人工智能带来的利与弊,不断强化政策分析与前沿战略研究,加速和谋划大气科学的新发展趋势。With the remarkable progress of computer technology,satellite remote sensing and communication networks,numerical weather prediction(NWP)has been born from a theoretical conception 100 years ago to reality and developed continuously.As two revolutions in NWP,data assimilation and ensemble prediction have not only benefited from the technological improvements of NWP,but also promoted the century-long development of NWP both from the technical and from the conceptual point of views.Looking back to the past,with technological innovation,the statistical and empirical weather mapping method could not meet the requirements of modern weather forecasting,NWP came into being.Nowadays,the new paradigm of artificial intelligence(AI)for science is showing noteworthy advantages.It has led to unprecedented progress in nowcasting,short-range forecasting,climate projection,and large meteorological models.Looking into the future,we should actively embrace technological changes,objectively view the pros and cons of AI,and continuously strengthen policy analysis and frontier strategic research,to pursue the new development trend of atmospheric sciences.

关 键 词：数值天气预报 资料同化 集合预报 技术革新 人工智能 

分 类 号：P4[天文地球—大气科学及气象学]