Font Size: a A A

Source Apportionment Of PM10 Used By Hybrid Model In Chongqing

Posted on:2009-10-23Degree:MasterType:Thesis
Country:ChinaCandidate:L LuFull Text:PDF
GTID:2121360272474718Subject:Environmental Science
Abstract/Summary:PDF Full Text Request
Inhalable particulate matter (PM10) is an important index for atmospheric environmental quality; is also harmful for human health. So identifying the sources of PM10 and calculating the contribution rate of PM10, namely conducting the source apportionment of PM10, is of great significance for the establishment of the practical measures to control PM10.The source apportionment of PM10 was conducted in Chongqing for providing scientific basis and technical support for the preventing dust work of Chongqing. In the process, factor analysis (FA) model was used to identify the sources of PM10, and the hybrid model, which was created with the mix of chemical mass balance (CMB) model and genetic algorithm (GA) model, is used for souce apportionment to provide some references for the technical methods. The main conclusions are as follows:(1) The mass concentrations of PM10 in Chongqing were characteristic of seasonal and regional variety. The concentration of PM10 every sampling site in winter was higher than in summer. The order of concentrations from high to low was: Intergrated Industrial Parks, Sowntowns, Residential Areas and Tourist Districts.(2) The component mass concentrations of PM10 were characteristic of seasonal and regional variety. The component concentrations of PM10 in winter were higher than summer, especially the increase of Si, S and Al in inorganic elements, organic carbon and SO42-, NO3- and NH4+ in water-solubility ions were more obvious. In the developed areas that have lots of industries, commerces, population and traffics, the component concentrations of PM10 were generally higher than in other areas.(3) The average value of OC/EC was 5.53, which in winter was higher than in summer. This showed that the pollution of secondary organic carbon (SOC) was serious. The concentration of SOC was 10.56μg/m3 and 59.87% of OC, which was calculated by the minimum method. The value of NO3-/SO42- was low, and it showed that the water-solubility ions mainly came from stationary pollution source. There was good correlation among NH4+, NO3- and SO42-. According to the linear regression equations, the main components of the atmospheric small particles were NH4HSO4, (NH4)2SO4, NH4NO3, and so on.(4) Secondary particles, fugitive dust, cement dust, steel dust and traffic dust the five pollution factors of PM10 in Chongqing were identified by FA model, and their relative contribution rates were 46.59%, 29.32%, 8.02%, 11.41% and 5.16%. At last the main pollution sources of PM10 in Chongqing were identified as fugitive dust (including soil dust and road dust), cement dust, coal smoke dust, mobile exhaust gas dust, steel dust and secondary particles (including secondary sulfate, secondary nitrate and secondary organic carbon) to perfect the pollution sources classification system of PM10.(5) According to the practical database operation and the test used by CMB model, the hybrid model had good accuracy, and could solve the problem of excessive source collinearity to a certain extent with its analytical results reliable.(6) The contribution rates of secondary particles, road dust, cement dust, soil dust, coal smoke dust, mobile exhaust gas dust and steel dust in Chongqing were 34.45%,23.41%,10.74%,10.44%,8.45%,7.94%和6.12% calculated by the hybrid model. It showed that the pollution of PM10 was mainly affected by secondary particulates, fugitive dust and cement dust with the sum of their contribution rates 79.04%. Most of secondary come from the atmospheric secondary reaction conducted by emissions of coal smoke dust and mobile exhaust gas dust, so the PM10 provention and control work should focus on the control of fugitive dust, construction dust, burning coal dust and mobile mobile exhaust gas dust.
Keywords/Search Tags:Source Apportionment, Factor Analysis, Chemical Mass Balance, Genetic Algorithm, Hybrid Model
PDF Full Text Request
Related items