Spatio-temporal Patterns And Uncertainty Analysis Of Satellite Retrieved Atmospheric CO₂ Columns In China

Atmospheric CO₂ is the most important greenhouse gas in atmosphere.Atmospheric CO₂ concentration has been increasing year by year since the Industrial Revolution began,which greatly affects global climatic change.Satellite-based column-averaged dry air mole fraction of CO₂（XCO₂）provides atmospheric CO₂concentration data on a large geographic scale,which is greatly supportive for study in change of CO₂ concentarion in atmosphere.However,the following two main problems exist in satellite-based XCO₂:（1）the precision of satellite retrieved XCO₂ is influenced by surface state,observed atmospheric condition et al.,which leads to complicated uncertainty in XCO₂ data.As a result,large uncertainty exists in spatio-temporal patterns of satellite-based XCO₂ all over the globe,especially in China;（2）Mutiple XCO₂ data products are released from different retrieval algorithms,but a general criterion in assessing their performance is still not available yet.Aiming at the above two,focusing on the special area,China,this study is carried out from the following two:（1）use atmospheric chemical transport model GEOS-Chem,particularly the high-resolution nested model in East Asia and introduce the Chinese High Resolution Emission Gridded Data（CHRED）to GEOS-Chem simulations of CO₂ over the Chinese mainland as anthropogenic CO₂ emissions data.In this way,XCO₂ are simulated in a more finer scale than ever.（2）Build a general criterion for assessing the performance of individual XCO₂ products from different retrieval algorithms,which take full consideration of CO₂ characterisitcs and spatio-temporal patterns with intercomparison among model simulation XCO₂ and different satellite-based XCO₂ products.The detailed study is as bellow:Model simulations of CO₂ at high spatio-temporal resolution and the finest scales ever are obtained in China.This data set is then used to assess spatial uncertainty in satellite retrieved XCO₂ in China.Additionally,this study chooses typical regions and pays attention to analyze spatio-temporal uncertainty in satellite retrieved XCO₂ using XCO₂ data sets from five GOSAT retrieval algorithms（ACOS,NIES,OCFP,SRFP and EMMA）as well as GEOS-Chem simulation XCO₂.After intercomparisons among model simulation XCO₂ and different satellite-based XCO₂ products,the general performance of each individual satellite XCO₂ products as well as the regional error characteristics is finally acquired.Furthermore,this study discusses the most likely attribution affecting factors on retrieval precision,including aerosol and surface albedo,and reveals how regional uncertainty in satellite retrieved XCO₂ relates to them as well as the internal functioning mechanism of each retrieval algorithm.The study results indicate that:（1）The average levels of satellite retrieved XCO₂from long timeseries dataset in China reasonably reflect positive correlation of XCO₂with anthropogenic emissions by showing maximums in the north and negative correlation with strong absorption from vegetation by showing minimums in the northeast.The maximum region in the northwest is likely an illusion induced by high XCO₂ retrieval error attributed to the combined effect of aerosol and albedo in deserts.（2）XCO₂ retrievals from five algorithms demonstrate better agreement in eastern regions with strong anthropogenic emissions than those in western grids of desert with high brightness surface.（3）ACOS and SRFP are found to perform better than the other three algorithms（NIES,OCFP and EMMA）.（4）The uncertainty in satellite retrieved XCO₂ is likely to increase with AOD or albedo when both AOD and albedo are high.This study makes the following two major achievements:1.Spatio-temporal patterns of atmospheric CO₂ concentration are simulated in finer scales than ever.CHRED,which is generated from the investigation of surface emitting point sources that was conducted by the Ministry of Environmental Protection of China,is introduced to GEOS-Chem nested simulations of CO₂ over China.As a result,this study obtains a finer CO₂ concentration simulation data set of high spatio-temporal resolution.2.Reveal the spatial pattern of uncertainty in GOSAT retrieved XCO₂ in the typical study area,which enlightens the application of GOSAT retrieved XCO₂ data in future.In the latitude zone of 40°N in China,the discrepancy in the multiple GOSAT retrieved XCO₂ products is the smallest in the east,where the megacity of Beijing is located and where there are strong anthropogenic CO₂ emissions,which implies that XCO₂ from satellite observations could be reliably applied in the assessment of atmospheric CO₂ enhancements induced by anthropogenic CO₂ emissions.The large inconsistency among the multiple GOSAT retrieved XCO₂ products presented in western deserts with a high albedo and dust aerosols,moreover,demonstrates that further improvement in GOSAT retrieved algorithms is still necessary in such regions and more attention should be paid when using GOSAT retrieved XCO₂ data in this area.3.A general criterion about how to analyze and assess spatio-temporal uncertainty in satellite retrieved XCO₂ are brought up.In study area with typical characteristics of land cover,this method designs to comprehensively assess satellite retrieved XCO₂ from the following three aspects:（a）regional and spatio-temporal bias patterns between satellite retrieved XCO₂ and model simulation XCO₂;（b）regional and spatio-temporal bias patterns among XCO₂ data sets retrieved by different algorithms;（c）the reasonability assessment about displayed spatio-temporal patterns of different XCO₂ data sets,based on prior knowledge of atmospheric CO₂such as seasonal cycle and main influencing factors（anthropogenic emissions）.This study has proved that using inter-comparisons between multiple XCO₂ data sets,generated from satellite retrieval algorithms or model simulation,is effective in assess satellite retrieved XCO₂ especially when high precision ground based measurements are not available.