Skip to main navigation Skip to search Skip to main content

DOAS and solar spectroscopic measurements of atmospheric constituents

  • Mau Hing CHAN

    Student thesis: Doctoral Thesis

    Abstract

    This thesis describes the instrumentation development, atmospheric optical remote sensing, and inversion methods to retrieve tropospheric concentrations of nitrogen dioxide (NO2), water vapor and aerosols. Continuous concentration measurements of the atmospheric constituents were performed. The measurement results and the seasonal variations are reported and discussed. Most of the measurements were conducted in Hong Kong (22.338°N, 114.172°E), but some of the experiments were performed outside the territory, for example, Xinjiang (the northwestern region of China), the People's Republic of China. In order to conduct experiments to measure tropospheric NO2 concentration, two types of DOAS (Differential Optical Absorption Spectroscopy) instruments have been developed. With the application of an acousto-optic tunable filter (a solid-state wavelength tunable device for spectral light dispersion), the two types of DOAS instruments are configured into an active and a passive approach. In the active DOAS configuration, a xenon lamp acting as the active light source is used to probe the atmospheric NO2 molecules. In contrast, in the passive mode, instead of using an artificial light source, the broadband direct sunlight is used. In the retrieval of NO2 concentrations, both the residual light intensities from the active and passive approaches are analyzed by means of DOAS technique, which is an algorithm to separate molecular absorption from atmospheric scattering effect, including Rayleigh and Mie scattering factors. In the local atmosphere, using the solar DOAS technique, monthly mean NO2 and water vapor vertical column densities during the period October 2004 to June 2005 were 161009.5×-161044.7× molecules/cm2 (or equivalently, 3.89-5.69 µg/cm2) and 221035.4×-221098.9× molecules/cm2 (1.30-2.98 g/cm2), respectively. Solar DOAS experiments were also conducted during a trip in Xinjiang (38.4°N-48.7°N, 75.0°E-89.5°E). Among the selected sites in Xinjiang, the highest NO2 column density is Urumqi (43.807°N, 87.600°E), but the lowest column density appears in Lake Karaku (38.441°N, 75.046°E). The NO2 column densities were respectively measured to be 171041.1× and 15105.7× molecules/cm2. In addition, with the sunlight as the light source, the Langley method is applied to retrieve water vapor column density. The monthly mean water vapor column amount from January 2002 to April 2005 was determined to be 221042.6×-231038.1× molecules/cm2 (1.92-4.13 g/cm2). Two different sources of water vapor data, the MODIS precipitable water vapor (PWV) and the column water vapor density derived from the Hong Kong Observatory (HKO) relative humidity vertical profile, are used to correlate with the water vapor amount from the Langley method. The correlation studies find that the water vapor amounts from our measurements are proportional to that of from MODIS PWV (at 20.0°N, 115.0°E) and HKO relative humidity (at 22°18'47"N, 114°10'13"E), with linear correlation coefficients, R2, values of 0.262 and 0.660, respectively. Higher correlation coefficient is found for the water vapor amounts retrieved from the nearer HKO site. Aerosol features in the local atmosphere are also reported. The methodology is to employ a Sky Radiometer for the observation of direct and diffuse solar radiation. The SKYRAD inversion code is used to retrieve the aerosol features, including aerosol optical thickness (AOT), volume spectrum (or size distribution), and complex refractive index. In order to validate the AOT from Sky Radiometer, MODIS (Moderate Resolution Imaging Spectroradiometer) AOT at 550 nm is used to perform the correlation analysis. The MODIS optical thicknesses obtained from AQUA and TERRA correlate with the Sky Radiometer data with linear R2 values of 0.804 and 0.625, respectively. The routine aerosol observations have been performed for three and a quarter years. Dependence of the AOT on meteorological parameters is investigated. In general, the monthly mean ground-based AOT is inversely proportional to the monthly mean wind speed. Annual mean ground-based AOT from 2002 to 2004 at visible channels have increased by 33% (from 0.641 to 0.854) and 22% (from 0.565 to 0.687) for observations at 400 and 500 nm, respectively. In addition, the results of AOT and size distribution give evidence of an increasing trend of fine particles, in comparison to coarse particles. In the Angstrom analysis, fitting of the annual mean AOT values to the Angstrom’s equation reveals a significant 53% increase in the Angstrom exponent (from 0.7 to 1.07) over the three-year period from 2002-2004. In the study of yearly mean fine particle to coarse particle volume concentration ratio (PM2.5/PM10 ratio) during 2002-2004, the PM2.5/PM10 ratio has increased by 7.7% (from 0.558 to 0.601).
    Date of Award15 Feb 2007
    Original languageEnglish
    Awarding Institution
    • City University of Hong Kong
    SupervisorYuk Sun Andrew CHENG (Supervisor)

    Keywords

    • Atmospheric chemistry
    • Absorption spectra
    • Spectrum analysis
    • Measurement
    • Optical spectrometers

    Cite this

    '