题目：Accurate Medium-Range Global Weather Forecasting with 3D Deep Neural Networks

报告人：田奇

时间：2023年10月20日（周五），上午10:00—11:00

地点：31-904

报告摘要：

Weather forecasting is important for science and society. Currently, the most accurate forecast system is the numerical weather prediction (NWP) method, which represents atmospheric states as discretized grids and numerically solves partial differential equations (PDEs) that describe the transition between those states. However, this procedure is computationally expensive. Recently, AI-based weather forecasting methods have shown potential in accelerating weather forecasting by orders of magnitudes, but the forecast accuracy is still significantly lower than that of NWP methods. In this paper, we introduce an AI-based method for accurate, medium-range global weather forecasting. We show that 3D deep networks equipped with Earth-specific priors are effective at dealing with complex patterns in weather data, and that a hierarchical temporal aggregation strategy reduces accumulation errors in medium-range forecasting. Trained on 39 years of global data, our program, Pangu-Weather, is the first to obtain stronger deterministic forecast results on reanalysis data in all tested variables, when compared with the world’s best NWP system, the operational integrated forecasting system (IFS) of the European Centre for Medium-Range Weather Forecasts (ECMWF). Our method also works well with extreme weather forecasts and ensemble forecasts. When initialized with reanalysis data, the accuracy of tracking tropical cyclones is higher than ECMWF-HRES.

报告人简介：

田奇，华为云人工智能领域首席科学家，国际欧亚科学院院士，教育部长江讲座教授、国家自然科学基金海外杰青，中国科学院海外评审专家，IEEE/CAAI Fellow，中国人工智能学会吴文俊人工智能领域杰出贡献奖获得者。2018年6月-2020年3月担任华为诺亚方舟实验室计算视觉首席科学家。田奇教授高中毕业于成都七中（1987级），本科毕业于清华大学电子工程系（1992），后赴美国伊利诺伊大学香槟分校学习，师从Thomas S. Huang教授获博士学位（2002）。后历任美国德克萨斯大学圣安东尼奥分校计算机系助理教授、副教授、和正教授（2002-2019），2010年获Google Faculty Research Award, 2017年UTSA校长杰出研究奖、2016年获评多媒体领域10大最具影响力的学者，并于2018年入选国家级领军人才创新项目。田教授是IEEE TMM, TCSVT, TNNLS, ACM TOMM、Multimedia Systems Journal等多个期刊的Associate Editor。他拥有多项美国专利，在计算机视觉及多媒体方向顶级期刊及会议如IEEE TPAMI，IJCV，TIP，TMM，CVPR,  ICCV，ECCV，ACM MM上发表文章约750+余篇（包括220 篇IEEE/ACM期刊和210篇CCF A类会议文章），谷歌学术引用次数57000+，h指数为106，有8篇论文获最佳论文奖或者最佳学生论文包括ACM Multimedia等。田教授的盘古气象大模型论文发表于《Nature》杂志正刊。