Satellite observations of atmospheric carbon dioxide （CO2） provide a useful way to improve the understanding of global carbon cycling. In this paper, we present a comparison between simulated CO2 concentrations from an inversion model of the CarbonTracker Data Assimilation System （CTDAS） and satellite-based CO2 measurements of column-averaged dry air mole fraction （denoted XCO2） derived from version 3.3 Atmospheric CO2 Observations from Space retrievals of the Greenhouse Gases Observing SATellite （ACOS-GOSAT） L2 data products. We examine the differences of CTDAS and GOSAT to provide important guidance for the further investigation of CTDAS in order to quantify the corre- sponding flux estimates with satellite-based CO2 observations. We find that the mean point-by-point difference （CTDAS-GOSAT） between CTDAS and GOSAT XCO2 is -0.11 4-1.81 ppm, with a high agreement （correlation r = 0.77, P 〈 0.05） over the studied period. The latitudinal zonal variations of CTDAS and GOSAT are in general agreement with clear seasonal fluctuations. The major exception occurs in the zonal band of 0°-15°N where the difference is approximately 4 ppm, indicating that large uncertainty may exist in the assimilated CO2 for the low- latitude region of the Northem Hemisphere （NH）. Additionally, we find that the hemispherical/continental differences between CTDAS and GOSAT are typically less than 1 ppm, but obvious discrepancies occur in different hemispheres/continents, with high consistency （point-by-point correlation r = 0.79, P 〈 0.05） in the NH and a weak correlation （point-by-point correlation r = 0.65, P 〈 0.05） in the Southern Hemisphere. Overall, the difference of CTDAS and GOSAT is small, and the comparison of CTDAS and GOSAT will further instruct the inverse modeling of CO2 fluxes using GOSAT.
A regional chemical transport model,RAMS-CMAQ,was employed to assess the impacts of biosphere-atmosphere CO2 exchange on seasonal variations in atmospheric CO2 concentrations over East Asia.Simulated CO2 concentrations were compared with observations at 12 surface stations and the comparison showed they were generally in good agreement.Both observations and simulations suggested that surface CO2 over East Asia features a summertime trough due to biospheric absorption,while in some urban areas surface CO2 has a distinct summer peak,which could be attributed to the strong impact from anthropogenic emissions.Analysis of the model results indicated that biospheric fluxes and fossil-fuel emissions are comparably important in shaping spatial distributions of CO2 near the surface over East Asia.Biospheric flux plays an important role in the prevailing spatial pattern of CO2 enhancement and reduction on the synoptic scale due to the strong seasonality of biospheric CO2 flux.The elevation of CO2 levels by the biosphere during winter was found to be larger than 5 ppm in North China and Southeast China,and during summertime a significant depletion （≥ 7 ppm） occurred in most areas,except for the Indo-China Peninsula where positive bioflux values were found.
Atmospheric CO2 concentrations from January 2010 to December 2010 were simulated using the GEOS-Chem(Goddard Earth Observing System-Chemistry) model and the results were compared to satellite Gases Observing Satellite(GOSAT) and ground-based the Total Carbon Column Observing Network(TCCON) data. It was found that CO2 concentrations based on GOSAT satellite retrievals were generally higher than those simulated by GEOS-Chem. The differences over the land area in January and April ranged from 1 to 2 ppm, and there were major differences in June and August. At high latitudes in the Northern Hemisphere in June, as well as south of the Sahara, the difference was greater than 5 ppm. In the high latitudes of the Northern Hemisphere the model results were higher than the GOSAT retrievals, while in South America the satellite data were higher. The trend of the difference in the high latitudes of the Northern Hemisphere and the Saharan region in August was opposite to June. Maximum correlation coefficients were found in April, reaching 0.72, but were smaller in June and August. In January, the correlation coefficient was only 0.36. The comparisons between GEOS-Chem data and TCCON observations showed better results than the comparison between GEOS and GOSAT. The correlation coefficients ranged between 0.42(Darwin) and 0.92(Izana). Analysis of the results indicated that the inconsistency between satellite observations and model simulations depended on inversion errors caused by data inaccuracies of the model simulation's inputs, as well as the mismatch of satellite retrieval model input parameters.
The purpose of this paper is to provide a robust and flexible implementation of a proper orthogonal decomposition-based ensemble four-dimensional variational assimilation method(PODEn4DVar) through Rlocalization.With R-localization,the implementation of the local PODEn4DVar analysis can be coded for parallelization with enhanced assimilation precision.The feasibility and effectiveness of the PODEn4DVar local implementation with R-localization are demonstrated in a two-dimensional shallow-water equation model with simulated observations(OSSEs) in comparison with the original version of the PODEn4DVar with B-localization and that without localization.The performance of the PODEn4DVar with localization shows a significant improvement over the scheme with no localization,particularly under the imperfect model scenario.Moreover,the R-localization scheme is capable of outperforming the Blocalization case to a certain extent.Further,the assimilation experiments also demonstrate that PODEn4DVar with R-localization is most efficient due to its easy parallel implementation.
The purpose of this study is to describe an economical approach to an existing adaptive localization technique and its implementation in the proper orthogonal decomposition-based ensemble four-dimensional variational assimilation method(PODEn4DVar). Owing to the applications of the sparse processing and EOF decomposition techniques, the computational costs of this proposed sparse flow-adaptive moderation(SFAM) localization scheme are significantly reduced. The effectiveness of PODEn4 DVar with SFAM localization is demonstrated by using the Lorenz-96 model in comparison with the Smoothed ENsemble Correlations Raised to a Power(SENCORP) and static localization schemes, separately. The performance of PODEn4 DVar with SFAM localization shows a moderate improvement over the schemes with SENCORP and static localization, with low computational costs under the imperfect model.
