| Particulate matter (PM) is one of the main pollutants in the atmosphere, which is harmful to human. In which, PM10(atmospheric dynamics equivalent diameter less than10μm) and PM2.5(atmospheric dynamics equivalent diameter less than2.5um) are the main subject attracting more and more interest. It is very important to use the moderate resolution remote sensor on polar orbital satellite in PM researches, as it has many spectral bands with larger space coverage and shorter revisit cycle. However, there are still many key problems to be solved in the near future work. In this study, we focused on PM10, PM2.5and AOD (Aerosol Optical Depth) in Shanghai and its surrounding areas. Correlation between each two of them was analyzed. Meanwhile, enhanced PM10forecast model was built based on remote sensing derived AOD and corresponding meteorological parameters. What’s more, a dynamic air quality monitory scheme was built based on real time remote sensing data. This dynamic system combined the real time AOD from the direct broadcast polar orbital satellite remote sensing receiving system at East China Normal University and the meteorological fields from Global Forecast System (GFS), then drive air parcel trajectory model to forecast. The main conclusions can be summarized as follows:(1) PM10and PM2.5had significant correlation no matter data obtained from single monitoring station or averaged data from many stations in the city. PM10and PM2.5had higher concentration in winter and lower concentration in summer. The main sources of PM were the local emissions.(2) PM10and AOD had similar spatial distribution, both of which showed a decrease from north to south, but they had different seasonal characteristics. Suitable validation (Relative humidity validation or Boundary layer height validation) method should be considered to improve the correlation between PM10and AOD in different seasons. The multi-resolution analysis showed that PM10and AOD had the same periodicity. Meanwhile, a significant correlation was found between PM2.5and AOD. Thus, it is feasible to use satellite derived data for PM10and PM2.5monitoring.(3) Correlations between PM10and related meteorological factors were different at different cities as well as in different seasons, which was a little higher in the north. What’s more, correlations between PM10and related meteorological factors were improved when adding the satellite derived AOD as an important independent variable into the geographically weighted regression, however, the spatial non-stationary was also generated.(4) An enhanced PM10forecast model was built for different scale based on multi-stepwise regression and wavelet analysis. The correct ratio and accuracy rate obtained from the small scale model (based on the local city characteristics) were higher than the medium model (based on the regional). Satellite derived AOD was the most important variable for model construction.(5) Combining the direct broadcasted remote sensing data with the GFS meteorological fields, the dynamic trajectories of the inhalable particulate matter in near48hours were forecasted via air parcel trajectory model. In Shanghai, the sensitivity height for PM transportation from long distance was found at1500m.(6) The dynamic monitoring method was validated by a case study about heavy inhalable particulate matter pollution. The simulated results of dynamic system were basically identical to simulation outputs from HYSPLIT. Meanwhile, the vertical distributions about aerosol component retrieved from Calipso can also reflect the processes showed by dynamic monitoring.The innovations of this study can be summarized as follows:(1) A new method was built for quickly and dynamically monitoring the air quality with large spatial coverage during the research and development of the regional inhalable particulate matter dynamic monitoring system. The dynamic monitoring system was built based on a combination among the remote sensing data, the real-time retrieval algorithm, the meteorological forecasting fields and the air parcel trajectory model.(2) During PM10and AOD correlation analysis, we found that their correlation had a seasonal sensitivity to different validation methods (Relative humidity validation and Boundary layer height validation). Meanwhile, an enhanced PM10forecast method was built with long term accumulated PM10and AOD data. On the one hand, the accuracy of PM10forecast was improved by the enhancement. On the other hand, it also provides some reference for high precision PM2.5forecast.(3) With all around local source analysis, we also analyzed the pollution sources from long distance with the backward trajectory model. Especially, we calculated the contribution rate in the height of500m,1000m, and1500m from long distance transportation sources in quantitative. Finally, the most sensitive height for long distance transportation was found. |
PDF Full Text Request