Carbon Dioxide Inversion Method Under High Aerosol Conditions

Carbon dioxide（CO₂）is an important greenhouse gas in atmosphere.The warming caused by the greenhouse effect has caused frequent and severe extreme weather.Therefore,high-precision detection of CO₂ is a hot topic that is concerned by international community.Ground-based observation method is traditional CO₂ measurement method.This method lacks the ability to conduct global real-time detection in a wide range of areas.Satellite observations can cover the whole world and obtain stable and continuous data with good spatial and temporal consistency at the global and regional scales.It has great advantages and development prospects.The accuracy of global atmospheric CO₂ observation data is only 1%or more,so that it can be applied to global climate research better.Because of this,the accuracy of the targets currently proposed for atmospheric CO₂ detection is higher than 1%.For such high precision requirements,atmospheric conditions for CO₂ inversion are limited to fine weather with optical thickness of less than 0.3.There are still certain difficulties at present.In our country,the atmosphere is more turbid.In order to make the inversion precision reach such high accuracy,it is faced with greater difficulties.This paper takes the CO₂ inversion demand of China’s GF-5 satellite as its application background.,analyzes the basic principle of atmospheric CO₂ satellite remote sensing,and constructs the atmospheric CO₂ inversion model under high-value aerosol conditions.On this basis,the atmospheric CO₂ inversion method under high aerosol conditions was explored.Atmospheric CO₂ Satellite remote sensing obtains solar radiation through atmospheric and terrestrial Interactions.CO₂ inversion is the process of extracting CO₂ concentration information from other influence factors in the received signal.Under the condition of high aerosols,atmospheric itself and the complexity of the coupling with the ground represent a non-unique solution to radiative transfer equation.In the inversion,it appears as a result of unstable or non-normal convergence.CO₂ inversion accuracy is difficult to improve.The CO₂ inversion method is the basic technical feature to find the best result in the interval.In this paper,we use principal component analysis to provide initial value for subsequent inversion.In principal component analysis,the relationship between environmental parameters and inversion errors is analyzed,and various environmental parameters are classified so that the principal component analysis results are relatively accurate.In this paper,the CO₂ inversion method is a combination of principal component analysis method and photon probability density function PPDF method.This is a method thatconsiders the path of photons in the transmission and the probability of being absorbed from a statistical point of view.It is a method to study the relationship between spectral radiation transmission and atmospheric CO₂ concentration from the depth of photon transmission process.For the application of inversion method to optical thickness,based on the O_2A-band PPDF-D,this paper expands to the combination of CO₂ strong absorption band 2.0um to develop the PPDF-S inversion method,which uses 0.76um and 2.Oum jointly inverts the PPDF factor to improve the adaptability of the inversion factor to the environment,and finally achieves the purpose of inversion of CO₂ concentration using the 1.6um absorption band of CO₂.In order to verify the feasibility of the inversion method,the first verification is performed when the optical thickness is low.At this time,the Taklimakan Desert 2013 data is used for verification.The results show that the inversion provided by GOSAT is relatively small when the aerosol optical thickness is small.As a result,the deviations of the PPDF-D inversion results are about 0.5%.The inversion results obtained by the PCA and PPDF-D joint inversion are 0.38.%;at the same time its inversion variance is significantly lower than the PPDF-D inversion bias alone.Therefore,based on GOSAT-FTS data,the inversion accuracy of the algorithm can be controlled within 1%when the optical thickness is small.Secondly,when the optical thickness is larger,The results show that the use of principal component analysis and two-channel photon probability density function can effectively correct the scattering effects of clouds and aerosols when the optical thickness is small;when the optical thickness is large,the correction ability is reduced,but the 2.Oum channel is increased.The resulting three-channel photon probability density make up for the lack of the current two-channel photon probability density;it is compared with the TCCON data at the base station and inverted using GOSAT satellite data near the TCCON site from January 2016 to April 2016.The inversion results are compared with TCCON and the results show that the maximum error of inversion does not exceed 8 ppm.The preliminary experimental results show that the inversion method of this paper has higher accuracy.