利用生成对抗网络从地球静止卫星图像中反演大气运动矢量  

Retrieving Atmospheric Motion Vectors from Geostationary Satellite Images Using Generative Adversarial Networks

作　　者：李孝涌 陈科艺[1] Li Xiaoyong;Chen Keyi(School of Atmospheric Sciences,Chengdu University of Information Technology,Chengdu 610225,Sichuan,China;Network and Equipment Support Center,Taizhou Meteorological Bureau,Taizhou 318000,Zhejiang,China)

机构地区：[1]成都信息工程大学大气科学学院,四川成都610225 [2]台州市气象局网络与装备保障中心,浙江台州318000

出　　处：《激光与光电子学进展》2023年第17期108-116,共9页Laser & Optoelectronics Progress

基　　金：国家自然科学基金面上项目(41875039)。

摘　　要：深度学习技术中的生成对抗网络(GAN)方法由于能够从输入的大量数据中抽取变量特征,生成更为真实的气象图像预测,在遥感领域有较为广泛的应用。但目前该算法在大气运动矢量(AMVs)的反演中应用较少,而AMVs是数值天气预报资料同化系统所需的重要产品资料。基于此,本文提出了利用生成对抗网络pix2pix从静止气象卫星图像反演AMVs的方法,由pix2pix模型将遥感影像转换为200 hPa和850 hPa的矢量风场。在最佳的资料和模型架构条件下,该方法反演得到的AMVs与传统算法所得产品资料质量相当,且克服了传统算法高度订正困难、无法获得某一层面完整风场和低层样本数偏低等缺点。个例分析亦表明,该方法针对具体的天气系统也有良好的表现。Generative adversarial network(GAN),a deep learning technique,is widely applied in the field of remote sensing because of its ability to extract features from large input data and generate more realistic forecasts of meteorological images.At present,however,the application of GANs in atmospheric motion vector(AMV)retrieval is rare,although AMVs are important data source for numerical weather prediction(NWP),especially in data assimilation.Based on this,a method for retrieving AMVs from geostationary satellite images using pix2pix,a type of GAN,is proposed.The pix2pix model is used to convert remote sensing images into wind vector fields at 850 hPa and 200 hPa.With the best data and model architecture,the AMVs obtained by this method are comparable to the AMVs retrieved using traditional algorithms.This method avoids the drawbacks of traditional algorithms,such as the inability to obtain complete wind fields at a certain level,difficulty of height assignment,and sparse AMVs at lower atmospheric levels.Case analysis shows that this method also performs well for specific weather systems.

关 键 词：大气光学 气象学 大气运动矢量 深度学习 生成对抗网络 

分 类 号：P412.27[天文地球—大气科学及气象学]