FY-3G MWRI-RM观测亮温中条纹噪声的识别和去除  

Striping noise identification and mitigation for FY-3G MWRI-RM

作　　者：毛严 董慧杰[1] 邹晓蕾 MAO Yan;DONG Huijie;ZOU Xiaolei(School of Atmospheric Science/Joint Center of Data Assimilation for Research and Application,Nanjing University of Information Science&Technology,Nanjing 210044,China)

机构地区：[1]南京信息工程大学大气科学学院/资料同化研究和应用联合中心,南京210044

出　　处：《气象科学》2024年第5期949-960,共12页Journal of the Meteorological Sciences

基　　金：江苏省自然科学基金资助项目(BK20240714)。

摘　　要：FY-3G是我国首颗低倾角轨道降水卫星,于2023年4月16日成功发射升空,主要用于灾害性天气系统的降水监测,观测范围在50°S~50°N内。MWRI-RM(Microwave Radiation Imager for the Rainfall Mission)是FY-3G搭载的主载荷之一。与搭载在FY-3B/C/D上的微波成像仪不同,MWRI-RM设有26个通道,包括中心频率覆盖10~89 GHz的大气窗区通道1~10,54 GHz附近的温度探测通道11~18、118 GHz和166 GHz的弱降水探测通道19~23、以及183 GHz附近的湿度探测通道24~26。本文发现MWRI-RM 26个通道观测亮温均存在条纹噪声。频率范围在10.65~36.5 GHz观测资料中的条纹噪声相对较小,大部分在±0.3 K之间;频率范围在89~183 GHz观测资料中的条纹噪声较大,除大部分在±0.5 K之间外,有些可达±1 K。本文所使用的条纹噪声去除方法能使海陆边界和云边缘处的亮温突变不被误认为条纹噪声的一部分,保证亮温重构场的正确性。识别和消除条纹噪声是MWRI-RM亮温观测降水产品反演和数据同化应用的重要步骤。FY-3G is the first Chinese low-inclination-orbiting precipitation satellite,launched on April 16,2023.It aims at monitoring global precipitation within the latitudinal range of 50°S-50°N.The MWRI-RM(Microwave Radiation Imager for the Rainfall Mission)is a principal instrument onboard the FY-3G.Unlike previous microwave imagers on FY-3B/C/D,MWRI-RM has 26 channels,including atmospheric window channels 1-10 with frequencies ranging from 10 to 89 GHz,temperature sounding channels 11-18 around 54 GHz,light precipitation detection channels 19-22 located at 118 GHz and channel 23 at 165.5±0.75 GHz,and humidity sounding channels 24-26 near 183 GHz.This study reports on the striping noise present in MWRI-RM observations across all channels.The magnitude of striping noise in the brightness temperature observations for MWRI-RM channels with their frequencies below 36.5 GHz is the smallest(mostly within±0.3 K),while for other channels(89-183 GHz),it ranges from±0.5 to±1 K.The striping noise mitigation method employed in this study prevents sharp variations of brightness temperatures at land-sea boundaries and cloud edges from being erroneously identified as striping noise,thus avoiding any impact on reconstructed brightness temperature observations.The striping noise mitigation is a necessary step for precipitation retrieval from and data assimilation of MWRI-RM brightness temperature observations.

关 键 词：FY-3G MWRI-RM 条纹噪声 噪声去除 PCA/EEMD 

分 类 号：P407.7[天文地球—大气科学及气象学]