| CO2is one of the most important greenhouse gases causing global warming.The rapid increase of atmospheric CO2concentration has led to a series of problems such as sea level rise,frequent extreme weather and ocean acidification.China has became the largest carbon emitter in the world since 2006,facing tremendous pressure to reduce greenhouse gases emissions.Atmospheric chemical transport model is an effective tool to establish the relationship between surface carbon flux and atmospheric CO2 concentration.Therefore,using a global atmospheric chemical transport model to simulate atmospheric CO2and compare it with observational data is conducive to understand the temporal and spatial variation of CO2,and clarify the changes of carbon emissions and the status of carbon sources and sinks in terrestrial ecosystems in different regions of the world,which has an important significance in implementing policies of carbon emission reduction and coping with future climate changeThis research is divided into two parts.The first part is a comparative study of two different resolutions of the MOZART-4 model which is used for simulating atmospheric CO2 concentration,including contrastive analysis of spatial-temporal differences of atmospheric C02 simulated by the 1°×1° and 2.81°×2.86° resolution models in different regions of the world and different regions of China from 2009 to 2015.And based on the global CO2observation data of OBSPACK data set,the two resolution models were evaluated and compared by simulating the atmospheric CO2.The second part is a contrastive analysis of CO2 atmospheric column concentration(XCO2)between simulation and observation.The spatial distribution and temporal variation of the simulated XCO2in the different regions of world and different regions of China were analyzed in detail and compared against with the GOSAT retrievals.The uncertainties of the optimized surface carbon flux of CT2016 in different regions were discussed based on the biases between the simulated and retrieved XCO2.The main results are as follows1)The surface CO2concentration simulated by the 1°×1° resolution model is higher than that by the 2.81°x2.86°resolution model on the global scale.The difference between 1°×1° and 2.81°×2.86° resolution models in summer half year is larger than that in winter half year.And the difference between 1°×1° and 2.81°×2.86°resolution models of land area is large and uneven,while that of ocean area is small and even.There are significant differences in the consistency of surface CO2 concentrations between the two resolution models in different regions of the world.There are good consistencies in North America temperate zone,South America temperate zone and Eurasia temperate zone,and the differences in other regions are more obvious,especially in South American tropical zone,South Africa and Asian tropical zone,the simulated concentration of the 1°×1°resolution model in each month is significantly higher than the 2.81°x2.86°resolution model.The statistics of the 1°×1° resolution model are generally improved compared to the 2.81°÷2.86° resolution model.It has been greatly improved in North America,Antarctica and the oceans,with a certain degree of improvement in Africa,Australia,South America and Europe,and no significant improvement in Asia.The simulated surface CO2 concentration by the two resolution models is lower than the observation in general,and the 2.81°×2.86°resolution model was significantly lower.2)In China,the average annual and summer semi-annual simulated surface CO2 concentration by the 1°×11°resolution model is higher than that by the 2.81°×2.86°resolution model in general,and the winter semi-annual is slightly lower than the later.On national average,the simulated concentration by two resolution models have high consistency in spatial distribution,but with the exception of Northeast China,the differences between two resolution models were obvious in other regions,and the simulated concentration of 1°×1° resolution model was significantly higher than that of 2.81°×2.86° resolution model in East China and South China,and was significantly lower in Southwest China than 2.81°×2.86°resolution model3)The results show that the simulated XCO2and GOSAT retrieved XC02 have high consistency in spatial distribution.In general,the simulated XCO2in the northern hemisphere is lower than observation,especially in Eurasia cold zone and North Africa.The differences between simulation and observation in the southern hemisphere are relatively small,and the simulated XC02in Australia and South Africa is slightly higher than the observation.The average annual and winter semi-annual of the simulated XC02in China is lower than observation,especially in North China.Northwest China and Southwest China.And the differences between simulation and observation of the summer semi-annual are much smaller,and the simulated XC02in Southeast China and Loess Plateau area are obviously higher than the observation.The differences between simulated XCO2 and observation in different seasons and different regions of the world are significantly different.There are good consistency between simulation and observation in North American temperate zone and Eurasia temperate zone.However,the differences between simulation and observation in other regions are obvious,especially in South American temperate zone,South Africa,Asian tropical zone and Australia.The consistencies of simulation and observation in different regions of China are also significantly different.The difference in Southwest China is the most significant.Except for July to September of each year,simulated XCO2 is obviously lower than observation in the rest of the time.The second most significant difference between simulation and observation is in Southeast China,where the simulated XCO2 is significantly higher from July to September each year.4)The differences between simulated XCO2 and observation reflect the fact that there are problems with the surface carbon flux in a certain extent.In North America cold zone,South America tropical zone,Eurasia cold zone and Europe the carbon sinks were overvalued in summer;In South American temperate zone,South Africa,Asian tropical zone and Australian both the carbon sources and carbon sinks were undervalued;the carbon source in North American cold zone,the biomass carbon emissions in North Africa and the fossil fuel carbon emissions in Eurasia temperate zone were undervalued in winter;In North American temperate zone the carbon source was overvalued in winter.the overall carbon sink results of China's terrestrial ecosystems in CT2016 are more credible,but there is a large uncertainty in the CT-optimized carbon flux in Southern China.The carbon sink in Southeast China was obviously underestimated,while the carbon sink in Southwest China was obviously overestimated.In Northern China,except the Northeast China,in other regions the carbon sources were underestimated during the winter half year.Moreover,CT2016 products could not reflect the impact of extreme drought events on changes in carbon sinks. |
PDF Full Text Request