基于GOSAT卫星数据的大气甲烷反演误差分析及校正

大气甲烷(CH₄)高精度反演受到多种因素的影响，其中地表特征和大气状态的不确定性是重要的影响因素，如地表反射率、温度、湿度和压力廓线。地表反射率受到诸多因素的影响，难以获得精确的数据，会给反演结果带来较大误差。温度、湿度和压力廓线的不确定性亦是反演误差的重要来源，由此产生的系统误差难以避免，单独利用CH₄吸收带进行反演难以消除此种误差。针对各种参数不确定性的影响，本文提出比值光谱法和CO₂吸收带校正法进行校正。比值光谱法通过将绝对辐亮度谱转化为比值光谱，抑制地表反射率在反演过程中的作用。CO₂吸收带校正法利用CO₂ 1.61 μm吸收带，将CH₄柱含量转化为CH₄体积混合比，校正温度、湿度和压力廓线不确定性引起的系统误差。通过将两种校正方法结合，可同时抑制地表反射率和温度、湿度、压力廓线不确定性产生的影响，减小反演误差。利用温室气体观测卫星(GOSAT)的观测数据进行大气CH₄反演，采用比值光谱法和CO₂吸收带校正上述误差，结果显示校正后的CH₄体积混合比与GOSAT-Level2产品相当接近，反演精度可达-0.24%，反演结果较为稳健可靠。研究表明，比值光谱法和CO₂吸收带校正法可有效校正地表特征和大气状态参数不确定性引起的误差，提高大气CH₄反演精度。

Errors Analysis and Correction in Atmospheric Methane Retrieval Based on Greenhouse Gases Observing Satellite Data

High precision retrieval of atmospheric CH4 is influenced by a variety of factors .The uncertainties of ground proper‐ties and atmospheric conditions are important factors ,such as surface reflectance ,temperature profile ,humidity profile and pres‐sure profile .Surface reflectance is affected by many factors so that it is difficult to get the precise value .The uncertainty of sur‐face reflectance will cause large error to retrieval result .The uncertainties of temperature profile ,humidity profile and pressure profile are also important sources of retrieval error and they will cause unavoidable systematic error .This error is hard to elimi‐nate only using CH4 band .In this paper ,ratio spectrometry method and CO2 band correction method are proposed to reduce the error caused by these factors .Ratio spectrometry method can decrease the effect of surface reflectance in CH 4 retrieval by con‐verting absolute radiance spectrometry into ratio spectrometry .CO2 band correction method converts column amounts of CH4into column averaged mixing ratio by using CO2 1.61 μm band and it can correct the systematic error caused by temperature profile , humidity profile and pressure profile .The combination of these two correction methods will decrease the effect caused by surface reflectance ,temperature profile ,humidity profile and pressure profile at the same time and reduce the retrieval error .GOSAT data were used to retrieve atmospheric CH4 to test and validate the two correction methods .The results showed that CH4 column averaged mixing ratio retrieved after correction was close to GOSAT Level 2 product and the retrieval precision was up to-0.24% .The studies suggest that the error of CH4 retrieval caused by the uncertainties of ground properties and atmospheric conditions can be significantly reduced and the retrieval precision can be highly improved by using ratio spectrometry method and CO2 band correction method .