甲烷柱浓度红外高光谱遥感反演与验证

CH4 column retrievals from ground-and space-based infrared spectra and satellite validation

甲烷(CH_(4))浓度变化是当前气候变化研究中的一个焦点问题。红外高光谱遥感技术已经成为探测大气CH_(4)浓度变化的重要技术手段。本文针对红外高光谱CH_(4)地基遥感反演技术,介绍了国际上重要的观测网络,包括TCCON(Total Carbon Column Observing Network),NDACC-IRWG(Network for the Detection of Atmospheric Composition Change-the Infrared Working Group)和COCCON(COllaborative Carbon Column Observing Network),讨论了这些观测网络的主要特点,包括观测仪器、波段设置、反演算法、产品特性等。针对红外高光谱CH_(4)浓度卫星遥感反演,概述了国际上的CH_(4)卫星遥感发展现状。同时,以日本GOSAT(Greenhouse gases Observing SATellite)卫星为例,探讨了卫星CH_(4)遥感地基验证工作中的关键技术,阐明了地基遥感对于卫星CH_(4)遥感产品改进的重要性。最后,利用TCCON香河站的观测资料,对最新版的TROPOMI(TROPOspheric Monitoring Instrument)卫星CH_(4)观测反演数据产品进行了地基验证,表明在华北地区TROPOMI CH_(4)柱浓度产品达到了其观测精度目标设定要求;TROPOMI观测得到的CH_(4)柱浓度年增长率要略高于TCCON的观测结果,两者相差0.263±0.172%/a;地基与卫星的差值具有季节变化特征,春季卫星的观测值大于TCCON观测值约0.3%,秋季卫星的观测值小于TCCON观测值约0.2%。

Methane(CH_(4))is the second most important greenhouse gas in the Earth's atmosphere,after carbon dioxide(CO_(2)).Understanding the change in CH_(4)concentration is a challenging task in atmospheric research given that it has various sources.Remote sensing has now become an effective technique to monitor CH_(4)concentrations globally.In this study,we presented an overview of CH_(4)column retrievals based on ground-based Fourier Transform Infrared spectrometer(FTIR)and space-based infrared measurements.Satellite validations were also discussed.Currently,three ground-based remote sensing international observation networks provide CH_(4)columns:the Total Carbon Column Observing Network(TCCON),the NDACC-IRWG(Network for the Detection of Atmospheric Composition Change-the Infrared Working Group),and the COCCON(COllaborative Carbon Column Observing Network).The main characteristics of the three networks were presented and discussed in our study,such as the measurement instrument,the observed spectra,the retrieval algorithm,and the postcorrection.TCCON and COCCON provide dry-air column-averaged mole fraction of CH_(4)(XCH_(4))measurements,with a systematic/random uncertainty of 0.1/0.5%.NDACC provides a total column of CH_(4),with a slightly large systematic/random uncertainty of 0.2/1.0%.However,it also provides a vertical profile of CH_(4),which allows us to observe the CH_(4)variations in the troposphere and stratosphere separately.Regarding the satellite CH_(4)retrievals,we compared several popular sensors with a nadir-view geometry and their retrieval algorithms,such as the TANSO-FTS/GOSAT,TROPOMI/S5P,IASI/MetOp,and AIRS/Aqua.Basically,the short-wave infrared measurements(GOSAT and TROPOMI)have more sensitivity to the low troposphere,while the thermal infrared measurements(IASI and AIRS)are mainly sensitive to the mid-and upper troposphere.The difference in their vertical sensitivity comes from the CH_(4)-specific absorption lines in the infrared region.All satellite retrievals are affected by the cloud,aerosol,and surface parameters.They also need to be validated and calibrated against ground-based measurements.Here,key steps during the satellite CH_(4)validation were discussed,including the statistical parameters,the a priori substitution,the smoothing correction,and the surface altitude correction.Finally,we showed the CH_(4)retrievals observed by the ground-based FTIR system at Xianghe,North China.We operated TCCON-type and NDACC-type measurements for the Bruker 125HR instrument and COCCON-type measurements for the Bruker EM27/SUN instrument.The entire FTIR measurement system at Xianghe was well described.Then,we used the TCCON XCH_(4)measurements to validate the co-located TROPOMI satellite observations within 50 km at Xianghe.The mean difference between TCCON and TROPOMI XCH_(4)measurements from June 2018 to May 2021 is 0.109%(nearly 2 ppb),which is within the retrieval uncertainty of the TROPOMI measurement.Moreover,a high correlation(R=0.92)is found between TCCON and TROPOMI XCH_(4)measurements at Xianghe.However,the annual growth of XCH_(4)derived from the TROPOMI satellite measurements is 0.263%±0.172%/year larger than that derived from the TCCON measurements.Besides,seasonal variation is observed in the differences between TCCON and TROPOMI XCH_(4)measurements,and the differences are obvious when the surface albedo is less than 0.1.Therefore,further investigations are needed to improve the TROPOMI CH_(4)retrievals in North China.