| As a remote sensing spectroscopic method, ground-based passive scattered light DOAS can successfully measure the column density of trace gases in stratosphere and troposphere for the whole layer. It’s an effective ways of comprehensive monitoring and evaluating the regional atmospheric pollution. With the rapid increase of vehicles and the gradual transformation of energy-use in Shanghai, the gaseous pollutants and particulate matters are coexist in the atmosphere, charactered as the composite pollution. Therefore, the air quality can be serious and has drawn scientists’ attention. Here, the ground-based passive DOAS observations were performed and the retrievals of NO2and aerosol pollution in the urban area of Shanghai were studied, which gave new insights to the regional atmospheric composite pollution. In the thesis, the principle, with its system design, assembly and working, of the MAX-DOAS and the mobile zenith scattered light DOAS are introduced detailed. The sensitivity studies of AMF are carried out for multiple parameters. The inversion algorithms are explored for aerosol optical properties, aerosol pollution types and boundary layer height. Furthermore, the related results are applied to improve the accuracy of the retrieved troposheric NO2VCD. The ground-based measurements are compared with the satellite tropospheric NO2products, i.e. GOME-2and OMI. The NO2, NOx emission from central urban area of Shanhgai are remote sensed using the mobile DOAS technique. The main contents and conclusions are as follows:(1) Through the sensitivity analysis, the influencing factors of AMF are indentified, which indicates that the observation geometry (solar zenith angle, relative azimuth angle and elevation angle), surface albedo, calculated wavelength and aerosol load, as well as trace gases profile, impact the AMF simulation. Since the observation geometry can be acquired from the experiment and the albedo varies smaller for urban ground, the aerosol load and NO2profile are mainly responsible for the AMF results. In particular, different aerosol loads and parameterization approach could result in huge discrepancy in tropospheric NO2AMF simulation.(2) After parameterizing the aerosl extinction coefficient, the multidimensional aerosol-04AMF lookup table was set up. According to the O4DSCD of different elevation angles from the MAX-DOAS observation, the inversion algorithms for aerosol optical parameters are realized based on the IDL language platform, including aerosol extinction coefficient and boundary layer height. The retrieved boundary layer heights are consistent to the results of micro-pulse lidar measurements and simulation from NECP-GADS. Diurnal variation of boundary layer height expressed with the rapid uplift in the morning, arrived to the stable maximum between11:00-14:00and reduced again from16:00. The maximum of boundary layer height and its diurnal variation range are occurred in summer rather than other seasons.(3) High correlations between zenith O4and H2O absorption in the visible band can be found through linear regression analysis no matter what the weather conditions are. Closely related to the atmospheric water vapor content, the H2O DSCDs showed agreements with the relative humidity and dew point. Slope values of linear regression are the minimum in the summer, as well as its fluctuations. As the correlation slope abruptly increased during the heavy dust due to low moisture content and enhanced O4absorption caused by abundant suspended dry crustal particles, the slope dropped suddenly in the haze episode owing to the significant augment of H2O absorption. Thus, the much discrepant correlation patterns may be regarded as a characteristic signature for dust and haze events.(4) To improve the accuracy of tropospheric NO2VCD, the retrieved boundary layer height is applied to the tropospheric NO2AMF simulation. The diurnal variation of tropospheric NO2VCD peaked high value in the rush hour, while reached a low value at noontime. Also the daily NO2VCD at weekend is less than working days. These results show that the measurement site, Fudan Univeristy, was significantly affected by vehicles and other anthropic emissions. Through the tropospheric NO2VCD comparison between different satellite products and different horizon scale,100km and50km grid size are determined for GOME-2, OMI products respectively. The ratio of ground-based observed tropospheric NO2VCD and GOME-2/DLR, OMI/TEMI are2.36±1.45,1.60±0.73. The discrepancy was mainly due to large spatial variability in the distribution of pollution intensity within the satellite grid, and the ground-based observation site in a high polluted area. Additionally, the correlation coefficient with the OMI/TEMI product and its difference are both better than GOME-2. The correlation coefficients for all weather conditions and sunny conditions are0.85and0.95, respectively.(5) To encircle along the inner ring viaduct road and Expo park in Shanghai, the NO2VCD measured by mobile DOAS tenichque has good correlation with satellite products and the traditional in-situ measurements in the Expo park. The characteristics of NO2VCD temporal variation and spatial distribution indicated that vehicles are the main emission source in the central urban area, influenced by human activities, traffic flow, vehicle operating conditions, and the wind field conditions. Based on the principle of continuity equation, the applicable conditions to estimate regional pollutant emissions by mobile DOAS technique are defined. The unperfected days due to the exisitence of clouds and heavey aerosols are screened via the O4absoprtions. The averaged NO2emissions rates from the IRVR area were determined to be2.1±0.9ton/h,2.8±1.4ton/h and2.7±1.4ton/h for autumn in2009, spring and autumn in2010. The annual vehicle NOX emissions for the central urban area were estimated to be (2.3±1.0)×104ton in2009and (3.0±1.5)×104ton in2010, and those for the whole city were (13.4±5.9)×104ton in2009and (17.6±8.4)×104ton in2010. |
PDF Full Text Request