Observational And Modeling Analysis Of Black Carbon In China

The black carbon aerosol (BC) is widely regarded as the second most importantwarming component in the atmosphere after CO2. One limitation of BC research inChina is the lack of continuous measurement with high temporal resolution. To addressthis issue, research conducted in this thesis is organized as follows: First, we set upcontinuous BC observation at Miyun; Second, after considering atmospheric processes,we calculated the top-down BC/CO emission ratio from observations and analyzedregional emission characteristics represented by the Miyun site; Third, by comparingthe observations with simulations by a chemical transport model (CTM), we estimatedthe possible biases in the model; Finally, we evaluated the newest-available BCemission inventory for China using the model.BC mass concentrations were measured continuously at Miyun, a rural site nearBeijing which can reflect the regional characteristic of North China Plan (NCP). Duringthe observation period of April to October,2010, the mean BC concentration was2.26±2.33μg/m~3. In summer and early fall, about70%-100%air masses were from NCP,with the rest originating from continental background. In contrast, the contributions ofthese two groups are roughly equal in spring. The correlation slope between BC and CO(dBC/dCO) can be used to identify emission characteristics. Observed dBC/dCO was0.0050±0.0001μg m-3/ppbv for the regional air masses of NCP. After considering theatmospheric processes, we derived an average top-down BC/CO emission ratio of0.0095±0.0002μg m-3/ppbv, which was90%higher than dBC/dCO prior toatmospheric treatment. This value is21%lower than the average emission ratio fromthe recent bottom-up inventory of Zhang et al. This difference is most likely to beattributed by residential sector, which needs to have lower share in the total emissionsof BC and/or much lower BC/CO emission ratio.We used the GEOS-Chem atmospheric chemical model to simulate BC in China.By comparing the model results with observations from two continuous sites (Miyunand Chongming), we found that model biases were mainly from simulating highpollution events and wet removal process. After excluding some data to minimize thesebiases, the concentration ratio of model to measurement decreased from0.89to0.67atMiyun and1.1to0.73at Chongming, suggesting that the inventory was likely to underestimate BC emissions. The measurements at other three rural sites (Longfengshan,Taiyangshan and Gaolanshan) are used to evaluate the recent emission inventory overChina. The inventory appears to underestimate BC emissions of Center China, NCP andYangtze River Delta region by74%±74%,33%±22%and27%±19%respectively.But it may overestimate the emissions from Northeast China and Center-southregion by27%±71%and36%±76%respectively. The emission biases foundhere are within±40%, much lower than the uncertainty of the inventory itself,which is±208%. Our results are different from previous understandings that theBC emission in China has been underestimated by more than100%across thecountry. This suggests that it is important to use continuous observation withhigh temporal resolution instead of monthly-mean data to evaluate BC emissioninventory.