City University of Hong Kong

CityU Institutional Repository >
3_CityU Electronic Theses and Dissertations >
ETD - Dept. of Physics and Materials Science  >
AP - Doctor of Philosophy  >

Please use this identifier to cite or link to this item:

Title: Source attribution of particulate matter and gaseous pollutants in Hong Kong
Other Titles: Xianggang qu yu ke li wu ji qi tai wu ran wu yuan shu xing yan jiu
Authors: Cheung, Irene Yin Ping (張燕冰)
Department: Dept. of Physics and Materials Science
Degree: Doctor of Philosophy
Issue Date: 2005
Publisher: City University of Hong Kong
Subjects: Air -- Pollution -- China -- Hong Kong
Gases -- Environmental aspects -- China -- Hong Kong
Particles -- Environmental aspects -- China -- Hong Kong
Notes: CityU Call Number: TD884.5.C45 2005
Includes bibliographical references (leaves 205-216)
Thesis (Ph.D.)--City University of Hong Kong, 2005
xxii, 216 leaves : ill. (some col.) ; 30 cm.
Type: Thesis
Abstract: The present work aims to identify the sources of airborne particulate matter in Hong Kong and to estimate their contributions to the particulate mass concentrations by using nonparametric regression (NPR) and receptor modeling. NPR was employed to identify the relationship between the dependent and predictor variables without making any assumptions about the functional form of the relationship or the statistical distribution of the data. In this work, hourly concentrations of six air pollutants, namely sulphur dioxide (SO2), carbon monoxide (CO), respirable suspended particulates (RSP) (PM10), nitric oxide (NO), nitrogen dioxide (NO2) and nitrogen oxides (NOx) from the air monitoring stations at Tung Chung (TC), Central/Western (CW), Mong Kok (MK) and Tsuen Wan (TW) of the Hong Kong Environmental Protection Department (HKEPD) in the period of 2000-2003 were studied. These stations were representative of residential, mixed residential/commercial, urban roadside and mixed residential/commercial/industrial areas. The locations of the nearby sources were determined by NPR analyzes of the atmospheric concentrations of these pollutants on two variables, namely, wind direction and wind speed. It is generally believed that wind speed is useful for distinguishing ground level emissions from elevated emissions like those from aircrafts. The results indicated that the major source of air pollution in TC was the emission from jet fuel combustion. A major finding from the TC data was that SO2 could be used as a tracer for aircraft emissions at the Hong Kong International Airport (HKIA). The main air pollutants at CW and MK were concluded to originate from vehicular emission, and the ratio between NO and NOx could be used as tracers for vehicular emissions. In contrast, vehicular emissions and power-plant combustion were found significant contributors at TW. In order to provide more information and better understanding of the origin and impact of particulates, multivariate receptor modeling was performed on the PM10 data. Among the various receptor models, UNMIX was chosen in this study, which was a relatively new advanced technique and had several advantages over conventional factor analysis methods. The impacts of the sources were examined using source contributions estimated from UNMIX coupled with wind direction and wind speed and visualized using plots of nonparametric regression (NPR) and conditional probability function (CPF). PM10 in Hong Kong was sampled for 24 h once every 6 days. These data were used to examine the applicability of this new type of factor analysis as a source identification tool for Hong Kong. Before applying the method to the Hong Kong data, a daily PM2.5 simulated dataset developed by the U.S. Environmental Protection Agency (USEPA) was used for validation. The original dataset with daily sampling were first transformed into a dataset with sampling every 6 days (referred to as the transformed dataset). The source profiles derived from the original USEPA dataset and the transformed dataset were compared with the known source profiles (provided by USEPA). Seven and five source categories were identified for the original daily dataset and the transformed dataset, respectively. It was concluded that the model could generally identify the major sources and was expected to give better performances for larger datasets. After this validation, we devised the method to identify the sources of PM10 in Hong Kong. UNMIX was used to obtain four possible sources in Tung Chung (TC). They were crustal/soil materials (38%), fresh combustion and residual oil (33%), coal-fired, incineration and automobile (21%), and marine aerosols (8%), with the bracketed numbers representing the percentage contribution to the PM10 mass. The major sources for PM10 in TC were crustal/soil materials and fresh combustion/residual oil. Unmix was used to obtain five possible sources in Central/Western (CW). They were non-ferrous smelter/incineration/automobile emission (45%), marine aerosols (24%), crustal/soil materials (17%), residual oil/fresh combustion/secondary sulfate (13%) and coal-fired power/industrial emission (1%). The main contributor to PM10 found in CW was non-ferrous smelter/incineration/automobile emission. In Mong Kok (MK), five sources were resolved by UNMIX, including fresh combustion and street/road dust (34% by mass), coal-fired power/incineration and automobiles (26%), crustal/soil materials (23%), sea-salt aerosols (13%) and residual/heavy oil/ secondary aerosols (4%). Vehicular emission was the major contributor to PM10 at the roadside station of Mong Kok. UNMIX resolved six sources for PM10 in Tsuen Wan (TW), including coal-fired power/incineration/automobiles emission (28%), fuel-oil burning (27%), crustal/soil materials/road dust (22%), fresh combustion (18%), marine aerosols (3%) and industrial emission (2%). The major contributor to PM10 found in TW was coal-fired power/incineration/automobiles emission. In summary, agreement between the NPR results for gaseous pollutants and UNMIX results for PM10 was good. The impacts of sources on four air monitoring stations in Hong Kong were successfully examined using source contributions estimated from UNMIX coupled with wind directions and wind speeds and visualized using plots of NPR and CPF. The meteorological parameters obtained from the Hong Kong Observatory (HKO) can provide more information and better understanding of the source contribution. In general, both CPF and NPR plots on wind direction agreed well with the locations of suspected sources. CPF was a simple and fast method to identify the sources. In contrast, NPR on wind direction provided smoothed concentrations and associated uncertainties for different wind directions. A new method utilizing source contributions derived from UNMIX with NPR on wind speeds and directions as well as CPF on wind speed factors was also successfully applied to the ambient air monitoring station data in Hong Kong. Those plots clearly and accurately showed the directions and locations of possible sources. The results of this study are significant in helping understand the origins and mass contributions of aerosol sources in Hong Kong.
Online Catalog Link:
Appears in Collections:AP - Doctor of Philosophy

Files in This Item:

File Description SizeFormat
fulltext.html157 BHTMLView/Open
abstract.html157 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.


Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer