| In this study, after testifying an atmospheric transport model (GEOS-Chem), we use it to simulate global surface CO2concentration (2006～2010) and analyse its spatial and temporal distribution. Also, we discuss the possible reasons which lead to the seasonal variation in regions. Meanwhile, the global carbon fluxes inversion system is established which can be used to generate the optimized global carbon fluxes.In the beginning, we compare the CO2concentration simulated by GEOS-Chem (driven by GEOS-4and GEOS-5meteorological data), the outcome shows that model driven by GEOS-5can better simulate spatial and temporal pattern of global CO2. Furthermore, we use Globalview-CO2to testify the simulation ability of GEOS-Chem model. The results indicate that this model can reproduce the seasonal and interannual variations of observation sites during simulation period.Second, we analyse the surface CO2concentration in mid-latitude in the northern hemisphere based on global simulation results (2006-2010). It indicates that there exist two high CO2concentration centers locating in the east of Asia and the northeast of North America respectively. The interannual growth of these two regions is higher than others in the northern hemisphere. In former region, the maximum value of CO2concentration happens in spring while in latter region it occurs in winter. The time of maximum CO2concentration changes in each year. In both regions, the minimum values happen in summer. In addition, we analyze the character of CO2concentration in regions divided by different land cover. It indicates that the seasonal cycle amplitudes of forest, cropland, and grassland decrease in turn and these differences associate with regional plant biology characters and their meteorological environments factors (temperature and precipitation). In these regions, surface CO2concentration and leaf area index (LAI) have significant negative correlations. The results also indicate that surface CO2concentration has a significant negative correlation with temperature and precipitation in most part of land in the northern hemisphere. The relationship is closer in temperature-CO2correlation.The major plants are different in each region since their different regional climate which can be reflected by LAI. We take eastern China as an example to analyze difference of surface CO2in typical forests and croplands. The results show that amplitudes of surface CO2seasonal cycles in forest areas are larger than that of croplands and surface CO2concentration is smaller in northern forest than in southern forest. Correlation of LAI and surface CO2concentration performs as Daxing’an Mountains forests> Changbai Mountains forests> Jiangnan mixed forests> Huanan forests (Among forests); Dongbei Plain> Huabei Plain> Chengdu Plain> Jianghuai Plain (Among croplands). In addition, we find that there exist significant negative correlations between surface CO2concentration and air temperature, precipitation. Forests in Daxing’an Mountains and Chengdu Plain have strong CO2-air temperature relationships. Meanwhile, Huabei Plain and Forests in Daxing’an Mountains have closed CO2-precipitation relationships. Also, in some part of eastern China, surface CO2changes are one month later than the changes of precipitation.In the end, we establish global carbon fluxes inversion system (land+ocean) based on Bayesian inversion theory. It generates monthly mean carbon fluxes of each region from2007to2009.The inversed results indicate that biosphere land and ocean act as atmospheric carbon sinks and3years average global sink is-6.2Pg C/year. The largest sink locates in the north of Asia, Europe and North America. The amplitudes of carbon flux seasonal cycles are larger in land regions than in ocean regions. The seasonal variation is largest in mid-latitude of north hemisphere and the maximum value happens in boreal summer. |
PDF Full Text Request